﻿<s>
As_CSA previously_RR described_VVN ,_, anaccess_NN1 descriptorspecifies_VVZ the_AT allowed_JJ accesses_VVZ between_II a_AT1 module_NN1 and_CC a_AT1 range_NN1 ._. 
</s>
<s>
The_AT most_RGT likely_JJ explanation_NN1 for_IF this_DD1 is_VBZ that_DD1 LDA_NN1 maximises_VVZ the_AT between-class_JJ variation_NN1 and_CC not_XX the_AT data_NN '_NULL s_ZZ1 global_JJ variance_NN1 ._. 
</s>
<s>
Hence_RR ,_, it_PPH1 is_VBZ very_RG important_JJ that_CST we_PPIS2 supplement_VV0 the_AT black-box_JJ testing_NN1 by_II fully_RR exercising_VVG and_CC validating_VVG the_AT SDL_NP1 models_NN2 with_IW a_AT1 white-box_JJ coverage_NN1 testing_NN1 approach_NN1 &lsqb;_( 40_MC &rsqb;_) similar_JJ to_II that_DD1 typically_RR used_VMK to_TO test_VVI implementation_NN1 code_NN1 ._. 
</s>
<s>
Importantly_RR ,_, we_PPIS2 are_VBR able_JK to_TO evaluate_VVI optimizations_NN2 in_II combination_NN1 with_IW each_PPX221 other_PPX222 ._. 
</s>
<s>
Model_NN1 sensitivity_NN1 analysis_NN1 can_VM identify_VVI important_JJ data_NN sets_NN2 ,_, but_CCB does_VDZ not_XX assess_VVI the_AT uncertainty_NN1 in_II those_DD2 data_NN sets_NN2 ._. 
</s>
<s>
In_RR21 particular_RR22 ,_, MCTS_NP1 is_VBZ effective_JJ when_CS it_PPH1 is_VBZ difficult_JJ to_TO evaluate_VVI non-terminal_JJ states_NN2 so_CS21 that_CS22 traditional_JJ depth-limited_JJ search_NN1 methods_NN2 perform_VV0 poorly_RR ._. 
</s>
<s>
If_CS we_PPIS2 take_VV0 this_DD1 conditional_JJ distribution_NN1 as_II the_AT "_" reference_NN1 "_" distribution_NN1 ,_, we_PPIS2 can_VM use_VVI U.S._NP1 EPA_NP1 '_NULL s_ZZ1 recommendation_NN1 and_CC set_VVD our_APPGE DIN_NN1 criterion_NN1 at_II the_AT 75_MC percentile_NN1 (_( U.S._NP1 EPA_NP1 ,_, 2000_MC )_) ,_, which_DDQ is_VBZ 0.625_MC ?_? mg/L_NNU ._. 
</s>
<s>
Ohio_NP1 uses_VVZ an_AT1 ICI_NP1 standard_NN1 of_IO 46_MC and_CC EPT_NN1 of_IO 10_MC to_TO define_VVI a_AT1 water_NN1 as_CSA in_II compliance_NN1 ,_, which_DDQ resulted_VVD in_II a_AT1 conditional_JJ distribution_NN1 of_IO DIN_NN1 very_RG similar_JJ to_II its_APPGE marginal_JJ distribution_NN1 (_( dashed_JJ line_NN1 in_II Fig._NN1 ?_? 5_MC )_) ._. 
</s>
<s>
Such_DA an_AT1 adaptively_RR constructed_VVN function_NN1 can_VM be_VBI used_VVN to_TO build_VVI an_AT1 image_NN1 graph_NN1 ._. 
</s>
<s>
The_AT carrier_NN1 computes_VVZ the_AT individual_JJ scoresi(t)_NN1 as_II the_AT number_NN1 ofvi(tj)_VV0 that_CST are_VBR betweenand_NN1 ._. 
</s>
<s>
Thus_RR the_AT deterministic_JJ model_NN1 (_( 3.1_MC )_) with_IW the_AT objective_JJ function_NN1 and_CC certain_JJ constraints_NN2 modified_VVN as_CSA described_VVN above_RL ,_, plus_II constraints_NN2 (_( 4.3c_FO )_) ,_, yields_VVZ a_AT1 robust_JJ model_NN1 (_( RO_NP1 )_) for_IF water_NN1 supply_NN1 system_NN1 planning_VVG with_IW the_AT uncertainty_NN1 set_NN1 Z_ZZ1 for_IF facility_NN1 capacities_NN2 given_VVN in_II (_( 4.1_MC )_) ._. 
</s>
<s>
As_CSA shown_VVN on_II Fig._NN1 14_MC ,_, while_CS stability_NN1 fluctuates_VVZ ,_, the_AT concepts_NN2 remain_VV0 mostly_RR linked_VVN to_II the_AT homonymous_JJ ones_NN2 ._. 
</s>
<s>
In_BCL21 order_BCL22 to_TO achieve_VVI high_JJ coverage_NN1 ratio_NN1 ,_, Mon_NPD1 ._. 
</s>
<s>
In_II contrast_NN1 ,_, this_DD1 work_NN1 '_NULL s_ZZ1 modeling_NN1 is_VBZ based_VVN on_II Jacobi_JJ iteration_NN1 and_CC binary_JJ arithmetic_NN1 (_( provided_VVN by_II a_AT1 path_NN1 algebra_NN1 )_) ._. 
</s>
<s>
In_II a_AT1 numerical_JJ example_NN1 with_IW piecewise_JJ linear_JJ triangles_NN2 the_AT new_JJ method_NN1 displayed_VVD comparable_JJ computational_JJ efficiency_NN1 (_( error_NN1 vs._II number_NN1 of_IO degrees_NN2 of_IO freedom_NN1 )_) to_II that_DD1 of_IO a_AT1 conforming_VVG method_NN1 when_CS computing_VVG the_AT deformation_NN1 gradient_NN1 ._. 
</s>
<s>
Note_VV0 that_CST this_DD1 is_VBZ quite_RG different_JJ from_II classical_JJ epistemic_JJ logic_NN1 ,_, in_II the_AT sense_NN1 that_CST if_CS we_PPIS2 replace_VV0 O_ZZ1 by_II K_ZZ1 ,_, then_RT neither_DD1 of_IO these_DD2 sentences_NN2 is_VBZ valid_JJ ._. 
</s>
<s>
This_DD1 turn_NN1 to_II language_NN1 has_VHZ been_VBN a_AT1 feature_NN1 of_IO organisational_JJ research_NN1 in_II recent_JJ years_NNT2 (_( Oswick_NP1 et_RA21 al._RA22 ,_, 2000_MC ,_, Alvesson_NP1 and_CC Karreman_NP1 ,_, 2000_MC ,_, Hardy_NP1 ,_, 2001_MC and_CC Grant_NP1 et_RA21 al._RA22 ,_, 2005_MC )_) and_CC has_VHZ been_VBN used_VVN ,_, albeit_CS infrequently_RR ,_, in_II information_NN1 systems_NN2 research_NN1 (_( Heracleous_NP1 and_CC Barrett_NP1 ,_, 2001_MC )_) ._. 
</s>
<s>
A_AT1 different_JJ surrogate-assisted_JJ PSO_NN1 was_VBDZ proposed_VVN by_II Parno_NP1 et_RA21 al_RA22 ._. 
</s>
<s>
&lsqb;_( 19_MC &rsqb;_) where_RRQ a_AT1 DACE_NN surrogate_NN1 was_VBDZ used_VVN with_IW PSO_NN1 and_CC applied_VVN to_II a_AT1 6-D_NP1 groundwater_NN1 management_NN1 problem_NN1 and_CC to_II several_DA2 2-D_JJ test_NN1 problems_NN2 ._. 
</s>
<s>
The_AT Random_JJ scena_NN1 rio_NN1 ,_, was_VBDZ run_VVN three_MC times_NNT2 for_IF each_DD1 experiment_NN1 ,_, and_CC an_AT1 average_NN1 was_VBDZ taken_VVN ._. 
</s>
<s>
The_AT algorithm_NN1 evolves_VVZ architectures_NN2 and_CC connection_NN1 weights_NN2 simultaneously_RR ,_, each_DD1 individual_NN1 being_VBG a_AT1 fully_RR specified_VVD RBFNN_NP1 ._. 
</s>
<s>
These_DD2 inputs_NN2 and_CC outputs_NN2 are_VBR time_VV0 series_NN ,_, typically_RR at_II a_AT1 daily_JJ time_NNT1 step_NN1 ,_, and_CC extending_VVG for_IF many_DA2 months_NNT2 or_CC years_NNT2 ._. 
</s>
<s>
The_AT Wallaby_NN1 Creek_NN1 site_NN1 represents_VVZ wet_JJ temperate_JJ Eucalypt_NN1 forest_NN1 in_II ?_? the_AT Southeast_ND1 of_IO Australia_NP1 that_CST burned_VVD in_II a_AT1 fire_NN1 in_II 2009_MC ._. 
</s>
<s>
We_PPIS2 also_RR use_VV0 the_AT audio_JJ content_NN1 itself_PPX1 to_TO identify_VVI topics_NN2 for_IF these_DD2 programmes_NN2 ._. 
</s>
<s>
He_PPHS1 argues_VVZ that_CST the_AT impacts_NN2 associated_VVN to_II these_DD2 dynamic_JJ processes_NN2 can_VM be_VBI larger_JJR than_CSN those_DD2 shown_VVN in_II the_AT traditional_JJ assessments_NN2 of_IO the_AT costs_NN2 of_IO climate_NN1 change_VV0 that_CST have_VH0 been_VBN published_VVN ._. 
</s>
<s>
On_II the_AT other_JJ hand_NN1 ,_, he_PPHS1 could_VM also_RR request_VVI an_AT1 extra_JJ resource_NN1 more_RGR often_RR than_CSN others_NN2 ._. 
</s>
<s>
At_II runtime_NNT1 ,_, the_AT part_NN1 database_NN1 was_VBDZ used_VVN to_TO compute_VVI the_AT CPF_NP1 CPp_NP1 for_IF each_DD1 part_NN1 p_ZZ1 in_II the_AT two_MC input_NN1 shapes_NN2 to_TO be_VBI matched_VVN ._. 
</s>
<s>
Dan_NP1 Geer_NP1 had_VHD a_AT1 series_NN of_IO columns_NN2 in_II IEEE_NP1 Security_NN1 and_CC Privacy_NN1 magazine_NN1 whose_DDQGE content_NN1 mostly_RR fell_VVD into_II this_DD1 category_NN1 ,_, for_REX21 example_REX22 ,_, the_AT popular_JJ "_" The_AT Owned_JJ Price_NN1 Index_NN1 ,_, "_" which_DDQ calculated_VVD the_AT salable_JJ value_NN1 of_IO cybercrime_NN1 plunder_NN1 such_II21 as_II22 social_JJ security_NN1 numbers_NN2 (_( Geer_NP1 and_CC Conway_NP1 ,_, 2009_MC )_) ._. 
</s>
<s>
After_II applying_VVG mesh_NN1 adaptivity_NN1 this_DD1 should_VM be_VBI close_JJ to_II unity_NN1 everywhere_RL except_CS at_II domain_NN1 partition_NN1 boundaries_NN2 (_( where_CS mesh_NN1 adaption_NN1 was_VBDZ explicitly_RR disabled_JJ )_) ._. 
</s>
<s>
The_AT leitmotif_NN1 returns_VVZ more_DAR and_CC more_RGR strongly_RR as_CSA the_AT case_NN1 unfolds_VVZ ,_, although_CS on_II this_DD1 first_MD occasion_NN1 ,_, no_AT particular_JJ alarm_NN1 is_VBZ raised_VVN ._. 
</s>
<s>
Unfortunately_RR ,_, this_DD1 modification_NN1 alone_RR is_VBZ not_XX sufficient_JJ as_CSA it_PPH1 introduces_VVZ a_AT1 linear_JJ dependence_NN1 between_II the_AT spaces_NN2 for_IF the_AT weightings_NN2 @iM_FO and_CC f_ZZ1 that_CST renders_VVZ the_AT method_NN1 unstable_JJ ._. 
</s>
<s>
The_AT concept_NN1 oflight-treewas_NN2 the_AT multicast_NN1 scenario_NN1 ,_, which_DDQ is_VBZ analogous_JJ to_II the_AT lightpath_NN1 idea_NN1 used_VVN in_II the_AT context_NN1 of_IO unicast_NN1 traffic_NN1 ._. 
</s>
<s>
The_AT study_NN1 of_IO a_AT1 regionalized_JJ variable_NN1 concerns_VVZ with_IW at_RR21 least_RR22 two_MC aspects_NN2 :_: (_( 1_MC1 )_) the_AT domain_NN1 in_II which_DDQ the_AT regionalized_JJ variable_NN1 is_VBZ defined_VVN ,_, and_CC (_( 2_MC )_) the_AT time_NNT1 dependent_JJ support_NN1 for_IF which_DDQ each_DD1 sample_NN1 of_IO the_AT regionalized_JJ variable_NN1 is_VBZ defined_VVN ._. 
</s>
<s>
Brief_JJ descriptions_NN2 of_IO uncertainty_NN1 related_VVN to_II these_DD2 procedural_JJ categories_NN2 related_VVN to_II small_JJ watersheds_NN2 appear_VV0 subsequently_RR in_II this_DD1 manuscript_NN1 ;_; for_IF detailed_JJ descriptions_NN2 see_VV0 Harmel_NP1 et@al_FO ._. 
</s>
<s>
Let_VV0 T_ZZ1 T_ZZ1 be_VBI a_AT1 subset_NN1 of_IO tasks_NN2 ._. 
</s>
<s>
It_PPH1 is_VBZ not_XX surprising_JJ that_CST the_AT results_NN2 were_VBDR not_XX as_RG good_JJ in_II some_DD cases_NN2 using_VVG the_AT original_JJ LSI_NP1 queries_NN2 ;_; since_CS the_AT techniques_NN2 themselves_PPX2 differ_VV0 in_II how_RRQ they_PPHS2 build_VV0 the_AT search_NN1 space_NN1 ,_, one_PN1 would_VM expect_VVI to_TO form_VVI queries_NN2 differently_RR for_IF LDA_NN1 than_CSN one_PN1 would_VM for_IF LSI_NP1 ._. 
</s>
<s>
Within_II each_DD1 server_NN1 ,_, VMs_NP1 are_VBR considered_VVN in_II decreasing_JJ order_NN1 of_IO their_APPGE volume-to-size_JJ ratio_NN1 (_( VSR_NP1 )_) ;_; where_RRQ VSR_NP1 is_VBZ defined_VVN as_CSA Volume/Size_NN1 ;_; size_NN1 is_VBZ the_AT memory_NN1 footprint_NN1 of_IO the_AT VM_NP1 ._. 
</s>
<s>
However_RR ,_, we_PPIS2 can_VM also_RR derive_VVI a_AT1 Twixonomy_NN1 for_IF any_DD user_NN1 p∈P_FO employing_VVG the_AT same_DA method_NN1 as_CSA the_AT one_PN1 illustrated_VVD in_II Algorithm_NN1 1_MC1 ._. 
</s>
<s>
Nevertheless_RR ,_, when_CS the_AT data_NN is_VBZ projected_VVN into_II a_AT1 lower_JJR dimension_NN1 subspace_NN1 derived_VVN through_II PCA_NP1 or_CC LDA_NP1 this_DD1 performance_NN1 increases_VVZ between_II 63%_NNU and_CC 76%_NNU ._. 
</s>
<s>
Each_DD1 packet_NN1 sent_VVD was_VBDZ offset_VVN by_II at_RR21 least_RR22 one_MC1 second_NNT1 from_II the_AT previous_JJ send_VV0 to_TO prevent_VVI the_AT possibility_NN1 of_IO two_MC or_CC more_DAR nodes_NN2 transmitting_VVG simultaneously_RR ._. 
</s>
<s>
Certificates_NN2 are_VBR issued_VVN by_II the_AT Posta_NP1 CA_NP1 ,_, the_AT official_JJ certification_NN1 authority_NN1 (_( CA_NP1 )_) of_IO the_AT Slovenian_JJ post_NN1 ,_, which_DDQ meets_VVZ national_JJ and_CC European_JJ regulations_NN2 for_IF issuing_JJ QC_NN1 ._. 
</s>
<s>
As_II a_AT1 comparison_NN1 ,_, we_PPIS2 employed_VVD a_AT1 technique_NN1 which_DDQ is_VBZ very_RG close_JJ to_II our_APPGE framework_NN1 &lsqb;_( 33_MC &rsqb;_) ._. 
</s>
<s>
It_PPH1 runs_VVZ as_II a_AT1 fully_RR distributed_VVN ,_, shared-nothing_JJ parallel_JJ application_NN1 ,_, in_II which_DDQ sites_NN2 are_VBR carved_VVN up_RP and_CC assigned_VVN to_II machines_NN2 for_IF crawling_VVG ._. 
</s>
<s>
The_AT mobile_JJ device_NN1 would_VM play_VVI an_AT1 active_JJ role_NN1 in_II storing_VVG user_NN1 profile_NN1 information_NN1 ,_, aggregating_VVG and_CC brokerage_VV0 services_NN2 ._. 
</s>
<s>
Then_RT again_RT an_AT1 anomaly_NN1 score_NN1 can_VM be_VBI used_VVN together_RL with_IW a_AT1 threshold_NN1 to_TO identify_VVI anomalies_NN2 ._. 
</s>
<s>
Law_NN1 and_CC Callon_NP1 (_( 1988_MC )_) explain_VV0 that_CST a_AT1 negotiation_NN1 space_NN1 makes_VVZ it_PPH1 possible_JJ for_IF mistakes_NN2 to_TO occur_VVI in_RR21 private_RR22 ;_; within_II a_AT1 negotiation_NN1 space_NN1 ,_, it_PPH1 is_VBZ also_RR possible_JJ to_TO experiment_VVI and_CC ,_, if_CS all_DB goes_VVZ well_RR ,_, it_PPH1 is_VBZ possible_JJ to_TO create_VVI relatively_RR durable_JJ socio-technical_JJ combinations_NN2 ._. 
</s>
<s>
The_AT AHP_NN1 process_NN1 involves_VVZ pair-wise_RR comparisons_NN2 between_II alternative_JJ goals_NN2 to_TO obtain_VVI users_NN2 '_NULL relative_JJ ranking_NN1 of_IO their_APPGE importance_NN1 ._. 
</s>
<s>
Consumers_NN2 are_VBR modelled_VVN as_CSA participants_NN2 who_PNQS will_VM change_VVI their_APPGE own_DA behaviour_NN1 in_II31 response_II32 to_II33 financial_JJ and_CC comfort_VV0 considerations_NN2 ._. 
</s>
<s>
The_AT mapping_NN1 function_NN1 need_VM not_XX be_VBI explicitly_RR specified_VVN ._. 
</s>
<s>
Since_CS no_AT single_JJ empirical_JJ study_NN1 can_VM test_VVI every_AT1 facet_NN1 of_IO a_AT1 theory_NN1 ,_, the_AT validation_NN1 of_IO theoretical_JJ models_NN2 is_VBZ needed_VVN for_IF empirical_JJ tests_NN2 in_II multiple_JJ settings_NN2 to_TO determine_VVI if_CSW the_AT theory_NN1 is_VBZ incrementally_RR supported_VVN ._. 
</s>
<s>
We_PPIS2 believe_VV0 our_APPGE data_NN collection_NN1 methodology_NN1 of_IO using_VVG well-established_JJ secondary_JJ data_NN sources_NN2 and_CC cross-checking_VVG the_AT data_NN with_IW other_JJ prominent_JJ public_JJ sources_NN2 yielded_VVD a_AT1 fairly_RR representative_JJ collection_NN1 of_IO the_AT most_RGT important_JJ privacy_NN1 breach_NN1 incidents_NN2 in_II the_AT U.S._NP1 during_II the_AT stated_JJ time_NNT1 period_NN1 ._. 
</s>
<s>
We_PPIS2 provide_VV0 details_NN2 of_IO our_APPGE data_NN extraction_NN1 process_NN1 and_CC metrics_NN1 definitions_NN2 in_II the_AT following_JJ two_MC subsections_NN2 (_( Sections_NN2 3.1_MC and_CC .2_MC )_) ._. 
</s>
<s>
Businesses_NN2 use_VV0 the_AT Internet_NN1 to_TO transact_VVI commerce_NN1 with_IW consumers_NN2 and_CC other_JJ businesses_NN2 ._. 
</s>
<s>
Conversely_RR ,_, if_CS two_MC actions_NN2 are_VBR not_XX defined_VVN as_CSA conflicting_JJ then_RT they_PPHS2 are_VBR regarded_VVN as_CSA compatible_JJ ._. 
</s>
<s>
The_AT project_NN1 aimed_VVD to_TO re-invent_VVI the_AT way_NN1 of_IO doing_VDG networking_NN1 ,_, re-defining_VVG the_AT whole_JJ Internet_NN1 architecture_NN1 as_CS31 far_CS32 as_CS33 the_AT physical_JJ layer_NN1 ._. 
</s>
<s>
Isolated_JJ alerts_NN2 are_VBR first_MD correlated_VVN by_II using_VVG graph_NN1 techniques_NN2 ,_, which_DDQ result_VV0 in_II high-level_JJ correlation_NN1 results_NN2 ._. 
</s>
<s>
In_II our_APPGE model_NN1 ,_, the_AT network_NN1 supports_VVZ a_AT1 connection_NN1 between_II each_DD1 pair_NN of_IO distinct_JJ nodes_NN2 in_II V._NNU Throughout_II this_DD1 paper_NN1 we_PPIS2 assume_VV0 that_CST we_PPIS2 are_VBR estimating_VVG the_AT parameters_NN2 of_IO a_AT1 binomial_JJ distribution_NN1 ._. 
</s>
<s>
Uncertainty_NN1 costs_NN2 represent_VV0 intangible_JJ cognitive_JJ costs_NN2 that_CST reflect_VV0 risk_NN1 about_II future_JJ performance_NN1 levels_NN2 ._. 
</s>
<s>
For_REX21 instance_REX22 ,_, new_JJ releases_NN2 of_IO the_AT OpenMI_JJ interfaces_NN2 or_CC any_DD other_JJ component_NN1 interface_NN1 will_VM certainly_RR require_VVI a_AT1 tailoring_NN1 of_IO the_AT contained_JJ components_NN2 to_II updates_NN2 in_II interface_NN1 descriptions_NN2 such_II21 as_II22 changes_NN2 on_II attribute_NN1 names_NN2 or_CC deprecated_JJ method_NN1 names_NN2 ._. 
</s>
<s>
There_EX are_VBR multiple_JJ unofficial_JJ activities_NN2 converting_VVG official_JJ statistical_JJ data_NN to_TO LOD_VVI ,_, usually_RR using_VVG DCV_MC and_CC SKOS_NN2 ._. 
</s>
<s>
We_PPIS2 also_RR proposed_VVD a_AT1 stress-free_JJ RBC_NP1 model_NN1 which_DDQ leads_VVZ to_II membrane_NN1 properties_NN2 ,_, while_CS the_AT other_JJ RBC_NP1 models_NN2 suffer_VV0 from_II stress_NN1 anomalies_NN2 ,_, which_DDQ result_VV0 in_II triangulation-dependent_JJ deformation_NN1 response_NN1 and_CC an_AT1 anisotropic_JJ equilibrium_NN1 shape_NN1 ._. 
</s>
<s>
In_RR21 addition_RR22 ,_, it_PPH1 can_VM be_VBI shown_VVN thatmaxn∈N_FO |_NULL scope(n)_VV0 |_NULL (_( the_AT maximum_JJ routing_NN1 table_NN1 size_NN1 )_) is_VBZ minimized_JJ whenAssumption_NN1 ._. 
</s>
<s>
During_II this_DD1 time_NNT1 various_JJ models_NN2 of_IO IT_NN1 performance_NN1 have_VH0 been_VBN developed_VVN to_TO show_VVI that_DD1 IT_NN1 :_: impacts_NN2 organizational_JJ performance_NN1 via_II intermediate_JJ business_NN1 processes_NN2 (_( &lsqb;_( Davenport_NP1 ,_, 1993_MC &rsqb;_) and_CC &lsqb;_( Barua_NP1 et_RA21 al._RA22 ,_, 1995_MC &rsqb;_) )_) require_VV0 complementary_JJ organizational_JJ resources_NN2 such_II21 as_II22 workplace_NN1 practices_NN2 and_CC structures_NN2 (_( &lsqb;_( Powell_NP1 and_CC Dent-Medcalfe_NP1 ,_, 1997_MC &rsqb;_) and_CC &lsqb;_( Ray_NP1 et_RA21 al._RA22 ,_, 2005_MC &rsqb;_) )_) is_VBZ influenced_VVN by_II the_AT external_JJ environment_NN1 (_( Hunter_NP1 et_RA21 al._RA22 ,_, 2003_MC )_) and_CC as_II a_AT1 construct_NN1 ,_, should_VM be_VBI disaggregated_VVN into_II meaningful_JJ components_NN2 (_( Sethi_NP1 and_CC Carraher_NP1 ,_, 1992_MC )_) ._. 
</s>
<s>
Since_CS a_AT1 working_JJ path_NN1 can_VM not_XX use_VVI two_MC links_NN2 in_II opposite_JJ directions_NN2 on_II the_AT same_DA span_NN1 (_( or_CC edge_NN1 in_II the_AT graph_NN1 )_) ,_, then_RT two_MC connections_NN2 which_DDQ are_VBR protected_VVN together_RL can_VM not_XX use_VVI the_AT same_DA span_NN1 either_RR in_II the_AT same_DA ,_, or_CC opposite_JJ directions_NN2 ._. 
</s>
<s>
The_AT measure_NN1 uses_VVZ a_AT1 user_NN1 '_NULL s_ZZ1 tags_VVZ Tu_NP1 and_CC her_APPGE resources_NN2 Ru_NN1 as_CSA input_VVN respectively_RR :_: (_( 3_MC )_) H_ZZ1 (_( R_ZZ1 |_NULL T_ZZ1 )_) =_FO r∈Rt∈Tp_NN1 (_( r_ZZ1 ,_, t_ZZ1 )_) log2_NN1 (_( p_ZZ1 (_( r_ZZ1 |_NULL t_ZZ1 )_) )_) ._. 
</s>
<s>
It_PPH1 is_VBZ the_AT purpose_NN1 of_IO this_DD1 paper_NN1 to_TO initiate_VVI such_DA a_AT1 comparison_NN1 ._. 
</s>
<s>
Therefore_RR ,_, tensors_NN2 can_VM only_RR represent_VVI regular_JJ hypergraphs_NN2 ,_, and_CC are_VBR not_XX suited_VVN for_IF irregular_JJ hypergraphs_NN2 (_( i.e._REX hypergraphs_NN2 with_IW varying_JJ cardinalities_NN2 )_) ._. 
</s>
<s>
Any_DD device_NN1 access_NN1 to_II organizational_JJ information_NN1 or_CC assets_NN2 must_VM be_VBI removed_VVN ._. 
</s>
<s>
This_DD1 section_NN1 will_VM outline_VVI some_DD key_JJ facets_NN2 of_IO the_AT Assess_VV0 and_CC Monitor_VV0 phase_NN1 of_IO this_DD1 BYOD_NN1 Security_NN1 Framework_NN1 ._. 
</s>
<s>
The_AT outputs_NN2 of_IO the_AT interviewing_JJ process_NN1 are_VBR interview_NN1 transcriptions_NN2 ,_, and_CC interview_NN1 notes_NN2 which_DDQ can_VM be_VBI very_RG time_NNT1 consuming_VVG to_TO develop_VVI ._. 
</s>
<s>
There_EX are_VBR many_DA2 different_JJ finite_JJ difference_NN1 ,_, finite_JJ volume_NN1 and_CC finite_JJ element_NN1 methods_NN2 that_CST can_VM be_VBI used_VVN for_IF solving_VVG the_AT bidomain_NN1 equations_NN2 &lsqb;_( 6_MC &rsqb;_) ,_, &lsqb;_( 7_MC &rsqb;_) ,_, &lsqb;_( 8_MC &rsqb;_) ,_, &lsqb;_( 9_MC &rsqb;_) ,_, &lsqb;_( 10_MC &rsqb;_) ,_, &lsqb;_( 11_MC &rsqb;_) and&lsqb;12&rsqb;_FO ._. 
</s>
<s>
Therefore_RR ,_, the_AT model_NN1 could_VM not_XX be_VBI calibrated_VVN to_TO reproduce_VVI historical_JJ observations_NN2 and_CC its_APPGE accuracy_NN1 in_II this_DD1 regard_NN1 could_VM not_XX be_VBI tested_VVN ._. 
</s>
<s>
Up_II21 to_II22 now_RT ,_, we_PPIS2 have_VH0 assumed_VVN that_CST players_NN2 "_" understand_VV0 "_" each_PPX221 other_PPX222 '_NULL s_ZZ1 probability_NN1 spaces_NN2 ._. 
</s>
<s>
These_DD2 rules_NN2 are_VBR derived_VVN from_II the_AT Borda_NP1 rule_NN1 ._. 
</s>
<s>
It_PPH1 is_VBZ this_DD1 stability_NN1 and_CC standardisation_NN1 that_CST makes_VVZ the_AT test_NN1 collection_NN1 so_RG attractive_JJ ._. 
</s>
<s>
To_TO persuade_VVI people_NN that_CST there_RL '_NULL s_ZZ1 something_PN1 here_RL that_CST they_PPHS2 want_VV0 ,_, so_RR for_IF that_DD1 reason_NN1 I_PPIS1 decided_VVD that_CST we_PPIS2 needed_VVD to_TO go_VVI to_II the_AT big_JJ outside_JJ guys_NN2 we_PPIS2 '_NULL re_II also_RR trying_VVG to_TO link_VVI it_PPH1 though_RR to_TO make_VVI it_PPH1 seem_VVI properly_RR integrated_VVN into_II the_AT business_NN1 ._. 
</s>
<s>
Thus_RR ,_, the_AT measurement_NN1 takes_VVZ into_II account_NN1 the_AT similarity_NN1 between_II the_AT original_JJ contour_NN1 and_CC the_AT polygonal_JJ approximation_NN1 ._. 
</s>
<s>
In_II the_AT paper_NN1 of_IO Bossavit_NP1 and_CC Vérité_NN1 &lsqb;_( 21_MC &rsqb;_) an_AT1 alternative_JJ approach_NN1 to_II the_AT truncationN_NN1 ,_, which_DDQ allowsto_NN1 brought_VVD close_RR to_II ,_, or_CC if_CS suppJsC_NN1 ,_, on_II the_AT surface_NN1 of_IO the_AT conductorCitself_PPX1 ,_, is_VBZ employed_VVN ._. 
</s>
<s>
Agents_NN2 may_VM also_RR reason_VVI about_II how_RRQ to_TO perform_VVI tasks_NN2 more_RGR efficiently_RR ._. 
</s>
<s>
The_AT two_MC sequences_NN2 so_RR generated_VVN are_VBR then_RT each_DD1 forward_NN1 validated_VVD ._. 
</s>
<s>
The_AT problem_NN1 this_DD1 causes_VVZ is_VBZ best_RRT illustrated_VVN by_II an_AT1 example_NN1 ._. 
</s>
<s>
The_AT system_NN1 generates_VVZ kslots_NN2 at_II initialization_NN1 time_NNT1 with_IW size(Sli)_NN2 <_FO size_NN1 (_( Sli_NP1 +_FO 1_MC1 )_) ._. 
</s>
<s>
On_II the_AT contrary_NN1 ,_, both_DB2 the_AT moving_JJ background_NN1 and_CC alternative_JJ actor_NN1 video_NN1 sequences_NN2 emphasize_VV0 more_RRR on_II the_AT consistence_NN1 in_II their_APPGE temporal_JJ focus_NN1 of_IO attention_NN1 ._. 
</s>
<s>
The_AT dynamics_NN of_IO a_AT1 particular_JJ domain_NN1 are_VBR captured_VVN by_II a_AT1 set_NN1 of_IO sentences_NN2 called_VVN a_AT1 basic_JJ action_NN1 theory_NN1 ._. 
</s>
<s>
It_PPH1 can_VM be_VBI clearly_RR noticed_VVN that_DD1 iTCP/aware_FU modes_NN2 achieved_VVN low_JJ delays_NN2 and_CC high_JJ acceptance_NN1 percentages_NN2 while_CS the_AT unaware/classic_JJ modes_NN2 suffered_VVN from_II higher_JJR delays_NN2 and_CC lower_JJR acceptance_NN1 percentages_NN2 ._. 
</s>
<s>
Section_NN1 3_MC briefly_RR discusses_VVZ current_JJ approaches_NN2 to_II e-science_JJ experiment_NN1 validation_NN1 and_CC explains_VVZ why_RRQ it_PPH1 is_VBZ necessary_JJ to_TO perform_VVI validation_NN1 after_II an_AT1 experiment_NN1 was_VBDZ executed_VVN ._. 
</s>
<s>
Formal_JJ rigor_NN1 :_: As_II a_AT1 trade-off_NN1 to_II the_AT ease_NN1 of_IO comprehension_NN1 ,_, our_APPGE patterns_NN2 per_RR21 se_RR22 do_VD0 not_XX impose_VVI any_DD formal_JJ constraints_NN2 on_II resulting_VVG conceptual_JJ authorization_NN1 models_NN2 or_CC enforcement_NN1 architectures_NN2 ._. 
</s>
<s>
This_DD1 helps_VVZ the_AT proposed_JJ algorithm_NN1 to_TO classify_VVI these_DD2 gaps_NN2 as_CSA text_NN1 ._. 
</s>
<s>
Fig._NN1 8_MC shows_VVZ the_AT plots_NN2 of_IO the_AT mean_JJ entropy_NN1 values_NN2 of_IO the_AT cancer_NN1 and_CC control_NN1 groups_NN2 using_VVG MSE_NN1 ._. 
</s>
<s>
By_II comparison_NN1 and_CC analysis_NN1 ,_, we_PPIS2 explore_VV0 the_AT proposed_JJ ICRGMM_NP1 algorithm_NN1 as_II the_AT best_JJT GMM_NP1 configuration_NN1 ._. 
</s>
<s>
The_AT accuracy_NN1 of_IO the_AT used_JJ validation_NN1 scheme_NN1 and_CC contribution_NN1 with_IW a_AT1 correct_JJ formal_JJ model_NN1 are_VBR also_RR taken_VVN into_II account_NN1 ._. 
</s>
<s>
Without_IW any_DD hypotheses_NN2 or_CC causal_JJ propositions_NN2 ,_, chronologies_NN2 become_VV0 chronicles_NN2 ,_, but_CCB these_DD2 are_VBR still_RR valuable_JJ descriptive_JJ renditions_NN2 of_IO events_NN2 ._. 
</s>
<s>
Thus_RR the_AT solid_JJ stress_NN1 in_II the_AT normal_JJ direction_NN1 at_II any_DD height_NN1 is_VBZ equal_JJ to_II the_AT reduced_JJ weight_NN1 of_IO the_AT solid_JJ material_NN1 overburden_VV0 ._. 
</s>
<s>
We_PPIS2 also_RR direct_VV0 our_APPGE attention_NN1 to_II the_AT applicability_NN1 of_IO specific_JJ research_NN1 methods_NN2 in_II this_DD1 context_NN1 ._. 
</s>
<s>
In_RR21 addition_RR22 ,_, we_PPIS2 provide_VV0 the_AT final_JJ accuracy_NN1 achieved_VVN at_II the_AT end_NN1 of_IO the_AT teaching_NN1 session_NNT1 ,_, to_TO allow_VVI a_AT1 comparison_NN1 of_IO the_AT teaching_NN1 performance_NN1 beyond_II the_AT first_MD few_DA2 examples_NN2 shown_VVN on_II the_AT learning_NN1 curves_NN2 ._. 
</s>
<s>
There_EX have_VH0 been_VBN earlier_JJR investigations_NN2 of_IO the_AT use_NN1 of_IO FOL_NN1 provers_NN2 to_TO reason_VVI with_IW description_NN1 logics_NN2 ._. 
</s>
<s>
It_PPH1 is_VBZ envisaged_VVN that_CST this_DD1 activity_NN1 should_VM be_VBI operationalised_VVN by_II firstly_RR assembling_VVG a_AT1 group_NN1 of_IO appropriate_JJ personnel_NN2 –_- IT_NN1 ,_, information_NN1 security_NN1 ,_, strategic_JJ planning_NN1 and_CC commercial_JJ specialists_NN2 –_- to_TO review_VVI the_AT security_NN1 implications_NN2 of_IO the_AT strategy_NN1 ._. 
</s>
<s>
With_IW tie_NN1 breaking_VVG in_II31 favour_II32 of_II33 p_ZZ1 ,_, c1is_FO eliminated_VVN in_II the_AT first_MD round_NN1 ,_, so_CS the_AT score_NN1 vector_NN1 in_II the_AT second_MD round_NN1 is_VBZ 〈_NULL 4,3,2_MC 〉_NULL ,_, so_CS pis_NN2 eliminated_VVN in_II the_AT second_MD round_NN1 ._. 
</s>
<s>
As_CSA shown_VVN in_II the_AT loop_NN1 ,_, the_AT higher_JJR the_AT degree_NN1 of_IO customer_NN1 trust_NN1 ,_, the_AT higher_JJR the_AT "_" transition_NN1 to_II electronic_JJ payment_NN1 systems_NN2 "_" ._. 
</s>
<s>
Such_DA an_AT1 approach_NN1 is_VBZ clearly_RR not_XX practical_JJ using_VVG existing_JJ (_( non-biometric_JJ )_) techniques_NN2 ,_, such_II21 as_II22 PINs_NN2 or_CC passwords_NN2 ,_, due_II21 to_II22 their_APPGE inherently_RR intrusive_JJ operation_NN1 ._. 
</s>
<s>
In_II this_DD1 section_NN1 ,_, we_PPIS2 will_VM present_VVI a_AT1 series_NN of_IO theoretical_JJ results_NN2 allowing_VVG us_PPIO2 to_TO define_VVI symmetries_NN2 in_II the_AT antiunification_NN1 matrix_NN1 ,_, and_CC obtain_VV0 further_JJR time_NNT1 savings_NN2 ._. 
</s>
<s>
We_PPIS2 conclude_VV0 that_CST our_APPGE protocols_NN2 behaved_VVN as_CSA intended_VVN ._. 
</s>
<s>
Note_VV0 that_CST ,_, by_II displaying_VVG all_DB the_AT domain_NN1 relevant_JJ terms_NN2 ,_, we_PPIS2 offer_VV0 support_NN1 for_IF formulating_VVG the_AT users_NN2 query_VV0 in_II terms_NN2 that_CST are_VBR actually_RR used_VVN within_II the_AT service_NN1 collection_NN1 ._. 
</s>
<s>
In_RR21 addition_RR22 ,_, the_AT focus_NN1 on_II information_NN1 in_II ICN_NP1 allows_VVZ for_IF the_AT provision_NN1 of_IO ?_? information_NN1 ?_? security_NN1 by_II securing_VVG the_AT content_NN1 independently_RR from_II the_AT channel_NN1 over_II which_DDQ it_PPH1 is_VBZ transferred_VVN ._. 
</s>
<s>
In_II their_APPGE approach_NN1 ,_, they_PPHS2 use_VV0 a_AT1 storage_NN1 scheme_NN1 called_VVN aggregated_JJ deltas_NN2 which_DDQ associates_VVZ the_AT latest_JJT version_NN1 with_IW each_DD1 of_IO the_AT previous_JJ ones_NN2 ._. 
</s>
<s>
In_BCL21 order_BCL22 to_TO characterise_VVI the_AT state_NN1 of_IO the_AT network_NN1 below_II the_AT boundary_NN1 BijBij_NP1 ,_, let_VV0 there_EX be_VBI a_AT1 set_NN1 of_IO k_ZZ1 operational_JJ metrics_NN1 N=N1_FO ,_, N2_FO ,_, ,_, NkN=N1_FO ,_, N2_FO ,_, ,_, Nk_NP1 ._. 
</s>
<s>
Section_NN1 2_MC describes_VVZ some_DD of_IO the_AT problems_NN2 that_CST a_AT1 technology-alone_JJ solution_NN1 can_VM not_XX address_VVI ._. 
</s>
<s>
By_II preprocessing_VVG and_CC storing_VVG additional_JJ data_NN during_II ingestion_NN1 ,_, we_PPIS2 can_VM reduce_VVI lookup_NN1 times_NNT2 for_IF VM_NP1 ,_, DM_NNU and_CC VQ_NP1 queries_NN2 compared_VVN to_II CB_NP1 and_CC TB_NN1 approaches_NN2 ._. 
</s>
<s>
Each_DD1 individual_JJ feature_NN1 score_NN1 is_VBZ calculated_VVN by_II the_AT carrier_NN1 as_II the_AT likelihood_NN1 that_CST a_AT1 reported_JJ feature_NN1 value_NN1 belongs_VVZ to_II the_AT corresponding_JJ sample_NN1 distribution_NN1 recorded_VVN as_II part_NN1 of_IO the_AT user_NN1 profile_NN1 ._. 
</s>
<s>
But_CCB this_DD1 initial_JJ condition_NN1 will_VM guarantee_VVI us_PPIO2 that_DD1 is_VBZ trivial_JJ ,_, hence_RR ._. 
</s>
<s>
In_II this_DD1 situation_NN1 ,_, a_AT1 separate_JJ database_NN1 account_NN1 can_VM be_VBI used_VVN for_IF running_JJ unit_NN1 tests_NN2 ,_, and_CC for_IF that_DD1 account_NN1 ,_, a_AT1 policy_NN1 could_VM be_VBI specified_VVN to_TO append_VVI the_AT query_NN1 with_IW the_AT appropriate_JJ criteria_NN2 ._. 
</s>
<s>
This_DD1 collection_NN1 contains_VVZ 419_MC context_NN1 elements_NN2 composed_VVN of_IO an_AT1 average_JJ number_NN1 of_IO 32_MC words_NN2 and_CC 2.45_MC entities_NN2 each_DD1 ._. 
</s>
<s>
The_AT effectiveness_NN1 of_IO our_APPGE failure_NN1 recovery_NN1 protocol_NN1 is_VBZ evaluated_VVN in_II a_AT1 large_JJ number_NN1 of_IO simulation_NN1 experiments_NN2 ._. 
</s>
<s>
For_REX21 example_REX22 ,_, Fig._NN1 1_MC1 shows_VVZ groups_NN2 of_IO loosely_RR connected_JJ agent_NN1 subcommunities_NN2 which_DDQ could_VM potentially_RR execute_VVI search_NN1 in_II parallel_NN1 rather_II21 than_II22 sitting_VVG idle_JJ ._. 
</s>
<s>
Data_NN was_VBDZ collected_VVN for_IF each_DD1 search_NN1 conducted_VVN by_II the_AT user_NN1 ._. 
</s>
<s>
The_AT Simple_JJ Voltage_NN1 problem_NN1 ,_, with_IW the_AT always_RR constraint_NN1 compiled_VVN into_II an_AT1 envelope_NN1 action_NN1 ,_, but_CCB leaving_VVG the_AT TIFs_NP2 uncompiled_JJ ,_, is_VBZ in_II a_AT1 form_NN1 that_CST can_VM be_VBI tackled_VVN using_VVG popf2_FO (_( although_CS its_APPGE heuristic_JJ is_VBZ largely_RR blind_JJ to_II the_AT impact_NN1 of_IO TIFs_NP2 )_) ._. 
</s>
<s>
We_PPIS2 analysed_VVD the_AT data_NN using_VVG Bayesian_JJ Networks_NN2 (_( BNs_NP2 )_) as_CSA they_PPHS2 provide_VV0 an_AT1 ideal_JJ tool_NN1 for_IF undertaking_VVG risk_NN1 analysis_NN1 of_IO complex_JJ environmental_JJ and_CC social_JJ problems_NN2 such_II21 as_II22 this_DD1 (_( Barton_NP1 et_NP1 ?_? al._NNU ,_, 2008_MC ,_, Johnson_NP1 et_NP1 ?_? al._NNU ,_, 2010_MC ,_, Pollino_NP1 et_NP1 ?_? al._NNU ,_, 2007_MC ,_, Said_NP1 ,_, 2006_MC )_) ._. 
</s>
<s>
Submitted_VVN systems_NN2 consisted_VVD in_RP automatically_RR learned_VVN hypernymy_NN1 graphs_NN2 ,_, with_IW some_DD amount_NN1 of_IO noisy_JJ and_CC missing_JJ nodes_NN2 and_CC edges_NN2 with_II31 respect_II32 to_II33 the_AT three_MC gold_JJ standard_JJ taxonomies_NN2 ._. 
</s>
<s>
The_AT addresses_NN2 of_IO the_AT memory_NN1 regions_NN2 are_VBR obtained_VVN by_II issuing_JJ commands_NN2 to_II the_AT driver_NN1 ,_, which_DDQ in_II turn_NN1 performs_VVZ symbol_NN1 lookup_NN1 in_II the_AT Guest_NN1 kernel_NN1 ._. 
</s>
<s>
The_AT efficiency_NN1 of_IO the_AT key_JJ tree_NN1 approach_NN1 depends_VVZ critically_RR on_II whether_CSW the_AT key_JJ tree_NN1 remains_VVZ balanced_JJ ._. 
</s>
<s>
The_AT EREON_NN1 ontology_NN1 and_CC the_AT semantic_JJ rules_NN2 library_NN1 are_VBR the_AT specification_NN1 for_IF the_AT semantic_JJ services_NN2 described_VVN in_II Section_NN1 6_MC ._. 
</s>
<s>
Data_NN management_NN1 :_: This_DD1 module_NN1 presents_VVZ methods_NN2 for_IF storing_VVG and_CC manipulating_VVG large_JJ quantities_NN2 of_IO data_NN on_II a_AT1 computer_NN1 ._. 
</s>
<s>
The_AT current_JJ Internet_NN1 is_VBZ made_VVN up_RP of_IO thousands_NNO2 of_IO autonomous_JJ systems_NN2 that_CST coexist_VV0 ,_, cooperate_VV0 ,_, and_CC compete_VV0 for_IF usage_NN1 of_IO its_APPGE various_JJ resources_NN2 ._. 
</s>
<s>
On_II the_AT other_JJ hand_NN1 ,_, a_AT1 parameter_NN1 may_VM not_XX be_VBI bound_VVN if_CS there_EX is_VBZ an_AT1 unbound_JJ set_NN1 of_IO possible_JJ values_NN2 ._. 
</s>
<s>
These_DD2 studies_NN2 typically_RR also_RR solicit_VV0 expert_NN1 opinions_NN2 on_II probabilities_NN2 of_IO both_RR attack_VV0 and_CC defense_NN1 strength_NN1 ,_, and_CC combine_VV0 a_AT1 wide_JJ variety_NN1 of_IO measurements_NN2 in_II attempts_NN2 to_TO influence_VVI security_NN1 spending_NN1 decisions_NN2 ._. 
</s>
<s>
The_AT study_NN1 was_VBDZ conducted_VVN in_II an_AT1 industrial_JJ context_NN1 ,_, on_II 44_MC real-life_JJ systems_NN2 of_IO the_AT French-railway_JJ company_NN1 SNCF_NP1 ._. 
</s>
<s>
But_CCB these_DD2 are_VBR not_XX reasonable_JJ assumptionsand_NN1 as_CSA we_PPIS2 '_NULL ve_VV0 demonstrated_VVN ,_, relaxing_VVG them_PPHO2 even_RR a_RR21 little_RR22 yields_VVZ security_NN1 trouble_NN1 ._. 
</s>
<s>
Now_RT we_PPIS2 turn_VV0 to_TO examine_VVI the_AT effect_NN1 of_IO incremental_JJ limit_NN1 reductions_NN2 ._. 
</s>
<s>
One_MC1 of_IO the_AT key_JJ problems_NN2 in_II this_DD1 context_NN1 is_VBZ how_RRQ to_TO discriminate_VVI between_II operational_JJ overload_NN1 due_II21 to_II22 legitimate_JJ service_NN1 requests_NN2 as_CSA in_II a_AT1 flash_JJ crowd_NN1 or_CC malicious_JJ attacks_NN2 (_( e.g._REX a_AT1 DDoS_NP1 attack_NN1 )_) and_CC then_RT apply_VV0 adequate_JJ countermeasures_NN2 to_TO mitigate_VVI or_CC remediate_VVI the_AT problem_NN1 &lsqb;_( 19_MC &rsqb;_) ._. 
</s>
<s>
Instead_RR we_PPIS2 are_VBR arguing_VVG that_CST the_AT notions_NN2 underlying_VVG standard_JJ logics_NN2 are_VBR better_RRR suited_VVN to_II the_AT Semantic_JJ Web_NN1 than_CSN those_DD2 underlying_JJ Datalog_NN1 ._. 
</s>
<s>
However_RR ,_, from_II the_AT outset_NN1 in_II the_AT control_NN1 system_NN1 design_NN1 ,_, such_DA corrective_JJ action_NN1 is_VBZ usually_RR embedded_VVN ,_, so_CS21 that_CS22 the_AT closed-loop_JJ system_NN1 retains_VVZ the_AT required_JJ dynamic_JJ characteristics_NN2 and_CC automatically_RR follows_VVZ the_AT desired_JJ trajectory_NN1 correcting_VVG ,_, whenever_RRQV necessary_JJ ,_, the_AT effects_NN2 of_IO disturbances_NN2 that_CST would_VM otherwise_RR drive_VVI the_AT climate_NN1 system_NN1 away_II21 from_II22 this_DD1 trajectory_NN1 ._. 
</s>
<s>
The_AT set_NN1 U_ZZ1 is_VBZ shuffled_VVN and_CC a_AT1 fixed_JJ number_NN1 of_IO items_NN2 (_( e.g._REX ,_, 1000_MC )_) is_VBZ kept_VVN ,_, the_AT rest_NN1 discarded_VVD ._. 
</s>
<s>
The_AT beginning_NN1 of_IO each_DD1 session_NNT1 was_VBDZ dedicated_VVN to_II summary_NN1 review_NN1 of_IO the_AT individual_JJ interviews_NN2 ._. 
</s>
<s>
However_RR ,_, finding_VVG the_AT appropriate_JJ values_NN2 to_TO initialise_VVI the_AT models_NN2 '_NULL parameters_NN2 is_VBZ not_XX trivial_JJ &lsqb;_( 43_MC &rsqb;_) ._. 
</s>
<s>
At_II the_AT lower_JJR end_NN1 of_IO the_AT frequency_NN1 spectrum_NN1 ,_, basin-scale_JJ waves_NN2 (_( often-called_JJ seiches_NN2 )_) form_VV0 after_II the_AT thermocline_NN1 is_VBZ tilted_VVN in_II31 response_II32 to_II33 the_AT application_NN1 of_IO a_AT1 surface_NN1 wind_NN1 stress_NN1 ._. 
</s>
<s>
For_IF Yin_NP1 &lsqb;_( 77_MC &rsqb;_) ,_, any_DD research_NN1 strategy_NN1 could_VM in_II principle_NN1 be_VBI applied_VVN to_II the_AT different_JJ types_NN2 of_IO research_NN1 question_NN1 but_CCB certain_JJ research_NN1 strategies_NN2 are_VBR more_RRR suited_VVN to_II certain_JJ types_NN2 of_IO research_NN1 question_NN1 ._. 
</s>
<s>
Then_RT the_AT discrepancy_NN1 of_IO a_AT1 finite_JJ ±1_FO sequencex1_FO ,_, ,_, xnof_VV0 lengthnis_NN1 aminimal_JJ C_ZZ1 such_CS21 that_CS22 for_IF everythe_JJ subsequencexd_NN1 ,_, x2d_FO ,_, xn/ddisC_FU -bounded_NN1 ._. 
</s>
<s>
In_II other_JJ words_NN2 ,_, the_AT size_NN1 of_IO the_AT organisation_NN1 and_CC limited_JJ resources_NN2 are_VBR NOT_XX excuses_NN2 for_IF not_XX being_VBG forensically_RR ready_JJ ._. 
</s>
<s>
Here_RL ,_, the_AT "_" Make_VV0 Payment_NN1 "_" business_NN1 process_NN1 of_IO the_AT procurement_NN1 management_NN1 capability_NN1 is_VBZ modelled_VVN using_VVG the_AT BPMN_NP1 ._. 
</s>
<s>
This_DD1 subset_NN1 involves_VVZ variations_NN2 in_II facial_JJ expression_NN1 ,_, illumination_NN1 ,_, and_CC pose_NN1 ._. 
</s>
<s>
Each_DD1 dimension_NN1 includes_VVZ description_NN1 of_IO ways_NN2 that_CST CIO_NP1 '_NULL s_ZZ1 talked_VVN about_II institutionalizing_VVG that_DD1 dimension_NN1 ._. 
</s>
<s>
With_IW these_DD2 shortcomings_NN2 ,_, the_AT code_NN1 efficiency_NN1 is_VBZ very_RG difficult_JJ to_TO measure_VVI ._. 
</s>
<s>
Furthermore_RR ,_, our_APPGE ordering_NN1 of_IO protocols_NN2 is_VBZ strict_JJ :_: an_AT1 attack_NN1 that_CST succeeds_VVZ against_II a_AT1 strong_JJ security_NN1 protocol_NN1 ,_, will_VM also_RR succeed_VVI against_II the_AT weaker_JJR security_NN1 protocol_NN1 ._. 
</s>
<s>
Dynamic_JJ hydrological_JJ models_NN2 are_VBR typically_RR calibrated_VVN with_IW an_AT1 implicit_JJ emphasis_NN1 on_II streamflow_NN1 peaks_NN2 ,_, and_CC are_VBR therefore_RR particularly_RR sensitive_JJ to_II timing_NN1 mismatches_NN2 between_II rainfall_NN1 and_CC streamflow_NN1 ._. 
</s>
<s>
The_AT overhead_NN1 for_IF an_AT1 iteration_NN1 of_IO the_AT Sun–Wheeler_NP1 Gauss–Seidel_NP1 approach_NN1 is_VBZ less_DAR than_CSN that_DD1 of&lsqb;15&rsqb;_FO ,_, since_CS it_PPH1 uses_VVZ element_NN1 interface_NN1 connectivity_NN1 and_CC does_VDZ not_XX require_VVI solution_NN1 of_IO local_JJ systems_NN2 ._. 
</s>
<s>
The_AT situation_NN1 where_RRQ available_JJ information_NN1 is_VBZ ambiguous_JJ forms_NN2 an_AT1 exception_NN1 to_II this_DD1 view_NN1 :_: if_CS all_DB we_PPIS2 know_VV0 is_VBZ that_CST a_AT1 photo_NN1 was_VBDZ taken_VVN in_II Washington_NP1 ,_, it_PPH1 makes_VVZ sense_NN1 to_TO represent_VVI the_AT result_NN1 e.g._REX as_II the_AT union_NN1 of_IO Washington_NP1 DC_NP1 and_CC Washington_NP1 state_NN1 ._. 
</s>
<s>
To_TO summarise_VVI this_DD1 research_NN1 area_NN1 ,_, there_EX is_VBZ a_AT1 challenge_NN1 in_II applying_VVG the_AT established_JJ laws_NN2 of_IO armed_JJ conflict_NN1 to_II cyber_NN1 warfare_NN1 ._. 
</s>
<s>
This_DD1 value_NN1 should_VM not_XX be_VBI higher_JJR than_CSN 30%_NNU ,_, and_CC my_APPGE experiments_NN2 suggest_VV0 that_CST even_RR a_AT1 value_NN1 of_IO 10%_NNU might_VM be_VBI sufficient_JJ ._. 
</s>
<s>
The_AT ranges_NN2 of_IO variation_NN1 of_IO the_AT random_JJ input_NN1 (_( bending_VVG stiffnessD_NN1 )_) are_VBR also_RR indicated_VVN ._. 
</s>
<s>
We_PPIS2 consider_VV0 that_CST the_AT scope_NN1 of_IO an_AT1 APIM_NN1 should_VM not_XX only_RR include_VVI the_AT mechanism_NN1 itself_PPX1 but_CCB also_RR the_AT support_NN1 processes_NN2 surrounding_VVG the_AT management_NN1 of_IO APIMs_NP2 ._. 
</s>
<s>
Hence_RR ,_, the_AT pre-emphasized_JJ signal_NN1 y(n)_NN1 is_VBZ segmented_VVN into_II M_MC frames_NN2 of_IO 20_MC ms_NN1 duration_NN1 with_IW a_AT1 10_MC ms_NN1 overlap_NN1 between_II two_MC consecutive_JJ frames_NN2 to_TO retain_VVI a_AT1 good_JJ quality_NN1 of_IO the_AT signal_NN1 and_CC to_TO avoid_VVI loss_NN1 of_IO information_NN1 &lsqb;_( 4_MC &rsqb;_) ,_, &lsqb;_( 34_MC &rsqb;_) ._. 
</s>
<s>
The_AT authors_NN2 spent_VVD 60_MC man_NN1 hours_NNT2 evolving_VVG their_APPGE test_NN1 plan_NN1 methodology_NN1 and_CC creating_VVG their_APPGE test_NN1 plan_NN1 ._. 
</s>
<s>
The_AT basic_JJ idea_NN1 behind_II pruning_NN1 is_VBZ that_CST when_CS an_AT1 itemset_NN1 is_VBZ not_XX frequent_JJ in_II the_AT recent_JJ tilted-time_JJ windows_NN2 (_( it_PPH1 is_VBZ either_RR infrequent_JJ or_CC sub-frequent_JJ )_) ,_, the_AT corresponding_JJ windows_NN2 are_VBR dropped_VVN from_II the_AT tilted-time_JJ window_NN1 table_NN1 if_CS the_AT maximum_JJ error_NN1 rateis_NN1 met_VVD ._. 
</s>
<s>
The_AT following_JJ expression_NN1 was_VBDZ developed_VVN to_TO express_VVI End-Stuffing_NN1 in_II31 terms_II32 of_II33 packets_NN2 :_: (_( 7_MC )_) @_II The_AT observed_JJ variability_NN1 in_II the_AT audio_JJ file_NN1 '_NULL s_ZZ1 data_NN encoding_VVG stemmed_JJ from_II the_AT fact_NN1 that_CST characters_NN2 appeared_VVD to_TO be_VBI occasionally_RR modified_VVN in_II transmission_NN1 ,_, where_CS characters_NN2 received_VVD were_VBDR within_II 0x01_FO of_IO the_AT encoded_JJ hexadecimal_JJ value_NN1 in_II the_AT audio_JJ file_NN1 '_NULL s_ZZ1 Data_NN SubChunk_VV0 ._. 
</s>
<s>
Note_VV0 that_CST after_II choosing_VVG the_AT Professor_NN1 ,_, the_AT user_NN1 will_VM attempt_VVI to_TO establish_VVI the_AT lock_NN1 ,_, at_II the_AT risk_NN1 of_IO not_XX being_VBG able_JK to_TO do_VDI so_RR ._. 
</s>
<s>
They_PPHS2 diffuse_VV0 their_APPGE practices_NN2 ,_, tools_NN2 ,_, and_CC case_NN1 experience_NN1 through_II multiple_JJ engagements_NN2 ._. 
</s>
<s>
In_II this_DD1 situation_NN1 ,_, Eq_NN1 ._. 
</s>
<s>
(_( 11_MC )_) describes_VVZ the_AT cost_NN1 to_II the_AT modularity_NN1 value_NN1 of_IO a_AT1 node_NN1 i_MC1 choosing_NN1 to_TO leave_VVI X0_FO ._. 
</s>
<s>
Huff_VV0 '_NULL s_ZZ1 research_NN1 team_NN1 conducted_VVD interviews_NN2 of_IO 31_MC employees_NN2 from_II eight_MC different_JJ organizations_NN2 ._. 
</s>
<s>
Fig._NN1 9_MC shows_VVZ the_AT cost_NN1 per_II subscriber/Mbit_NN1 of_IO current_JJ and_CC next-generation_JJ EPON_NN1 and_CC WiMAX_VV0 networks_NN2 under_II triple-play_JJ traffic_NN1 for_IF three_MC different_JJ terrain_NN1 types_NN2 ._. 
</s>
<s>
Edited_VVN by_II A._NP1 Takeishi_NP1 ,_, B._NP1 Thorngren_NP1 and_CC S._NP1 Jarvenpaa_NP1 ._. 
</s>
<s>
Increasing_VVG the_AT alphabet_NN1 size_NN1 |_NULL |_NULL further_RRR does_VDZ not_XX improve_VVI the_AT results_NN2 and_CC increasing_VVG the_AT reduced_JJ state_NN1 length_NN1 m_ZZ1 creates_VVZ a_AT1 larger_JJR number_NN1 of_IO states_NN2 of_IO the_AT probabilistic_JJ finite_JJ state_NN1 automata_NN1 (_( PFSA_NP1 )_) ,_, many_DA2 of_IO them_PPHO2 having_VHG very_RG small_JJ or_CC near-zero_JJ probabilities_NN2 ._. 
</s>
<s>
Behind_II this_DD1 ,_, however_RR ,_, is_VBZ the_AT simple_JJ ontology_NN1 that_CST knows_VVZ that_CST Warsaw_NP1 is_VBZ a_AT1 city_NN1 in_II Poland_NP1 ,_, that_CST Poland_NP1 is_VBZ a_AT1 country_NN1 in_II Europe_NP1 ,_, etc_RA ._. 
</s>
<s>
Such_DA a_AT1 precondition_NN1 will_VM be_VBI a_AT1 large_JJ disjunction_NN1 that_CST includes_VVZ a_AT1 positive_JJ or_CC negative_JJ literal_JJ for_IF every_AT1 ground_NN1 atom_NN1 in_II the_AT planning_NN1 domain_NN1 ,_, and_CC the_AT number_NN1 of_IO ground_NN1 atoms_NN2 is_VBZ often_RR exponential_NN1 in_II the_AT size_NN1 of_IO the_AT domain_NN1 description_NN1 ._. 
</s>
<s>
In_II the_AT case_NN1 of_IO variable_JJ attributes_NN2 such_II21 as_II22 temperature_NN1 ,_, corresponding_VVG numerical_JJ data_NN will_VM be_VBI stored_VVN with_IW time-stamp_JJ information_NN1 in_II the_AT RTDB_NP1 ._. 
</s>
<s>
We_PPIS2 have_VH0 structured_VVN the_AT paper_NN1 by_II firstly_RR describing_VVG the_AT overall_JJ structure_NN1 and_CC content_NN1 of_IO our_APPGE source_NN1 material_NN1 ,_, the_AT definitive_JJ 90-volume_JJ edition_NN1 of_IO Leo_NP1 Tolstoy_NP1 '_NULL s_ZZ1 works_VVZ ._. 
</s>
<s>
Gathering_VVG this_DD1 user_NN1 feedback_NN1 makes_VVZ it_PPH1 possible_JJ to_TO automatically_RR refine_VVI the_AT automated_JJ algorithms_NN2 ._. 
</s>
<s>
As_CSA discussed_VVN earlier_RRR ,_, a_AT1 small_JJ value_NN1 means_VVZ that_CST few_DA2 valid_JJ symbols_NN2 will_VM be_VBI marked_VVN invalid_JJ but_CCB many_DA2 invalid_JJ ones_NN2 will_VM be_VBI marked_VVN valid_JJ ,_, whereas_CS a_AT1 large_JJ value_NN1 means_VVZ the_AT converse_NN1 ._. 
</s>
<s>
Agents_NN2 have_VH0 bounded_VVN rationality_NN1 and_CC gain_VV0 only_RR partial_JJ information_NN1 about_II their_APPGE environment_NN1 in_II local_JJ interactions_NN2 ._. 
</s>
<s>
When_CS considering_VVG the_AT computation_NN1 of_IO modularity_NN1 ,_, the_AT two_MC objectives_NN2 of_IO most_DAT community_NN1 detection_NN1 methods_NN2 are_VBR either_RR efficient_JJ running_JJ times_NNT2 or_CC performance_NN1 in_II obtaining_VVG higher_JJR value_NN1 of_IO the_AT objective_JJ function_NN1 given_VVN by_II modularity_NN1 ._. 
</s>
<s>
This_DD1 was_VBDZ counter_NN1 to_II findings_NN2 in_II other_JJ fuel_NN1 complexes_NN2 referenced_VVD above_RL ._. 
</s>
<s>
With_IW this_DD1 standardized_JJ information_NN1 ,_, the_AT clustering_NN1 algorithm_NN1 is_VBZ run_VVN and_CC the_AT permissions_NN2 are_VBR classified_VVN ._. 
</s>
<s>
However_RR ,_, as_CSA shown_VVN inTable_JJ 19_MC ,_, all_DB of_IO the_AT correlations_NN2 resulted_VVN inp-values>_FO ._. 
</s>
<s>
A_AT1 mixed_JJ metric_JJ constraint_NN1 is_VBZ a_AT1 comparison_NN1 of_IO numeric_JJ values_NN2 in_II which_DDQ both_RR controllable_JJ and_CC uncontrollable_JJ numeric_JJ fluents_NN2 appear_VV0 ._. 
</s>
<s>
Moreover_RR ,_, spatial_JJ integration_NN1 has_VHZ reduced_VVN the_AT FPR_NP1 of_IO the_AT ML_NNU detector_NN1 quite_RG dramatically_RR ,_, however_RR ,_, this_DD1 is_VBZ at_II the_AT expense_NN1 of_IO a_AT1 vast_JJ reduction_NN1 of_IO the_AT TPR_NP1 ._. 
</s>
<s>
In_BCL21 order_BCL22 to_TO solve_VVI this_DD1 problem_NN1 ,_, the_AT method_NN1 from_II &lsqb;_( 5_MC &rsqb;_) restricts_VVZ the_AT search_NN1 space_NN1 to_II a_AT1 dual_JJ band_NN1 of_IO predefined_JJ width_NN1 ._. 
</s>
<s>
In_II this_DD1 section_NN1 ,_, we_PPIS2 discuss_VV0 software_NN1 patterns_NN2 and_CC the_AT role_NN1 of_IO mapping_VVG studies_NN2 in_II software_NN1 engineering_NN1 as_CSA key_JJ background_NN1 concepts_NN2 ._. 
</s>
<s>
The_AT Inference_NN1 Inspector_NN1 approach_NN1 appears_VVZ to_TO work_VVI best_RRT when_CS knowledge_NN1 over_II an_AT1 existing_JJ signature_NN1 was_VBDZ changed_VVN (_( restrictions_NN2 or_CC definitions_NN2 on_II existing_JJ classes_NN2 )_) ._. 
</s>
<s>
This_DD1 is_VBZ why_RRQ the_AT number_NN1 of_IO tap_NN1 changes_NN2 remains_VVZ constant_JJ and_CC the_AT standard_JJ deviation_NN1 is_VBZ zero_MC ._. 
</s>
<s>
For_REX21 example_REX22 ,_, when_CS the_AT loss_NN1 run-lengths_NN2 of_IO two_MC different_JJ sequences_NN2 are_VBR characterised_VVN by_II different_JJ distributions_NN2 (_( e.g._REX ,_, one_PN1 showing_VVG Poisson-distributed_JJ loss_NN1 run-lengths_NN2 ,_, and_CC the_AT other_NN1 showing_VVG geometrically_RR distributed_VVN loss_NN1 run-lengths_NN2 )_) ,_, it_PPH1 is_VBZ unclear_JJ how_RRQ they_PPHS2 should_VM be_VBI compared_VVN ._. 
</s>
<s>
For_REX21 example_REX22 ,_, a_AT1 value_NN1 of_IO r=0.6_FO would_VM mean_VVI di_FW is_VBZ minimized_VVN over_II a_AT1 subset_NN1 of_IO 60%_NNU of_IO the_AT points_NN2 in_II Q_ZZ1 ,_, supposing_CS |_NULL Q_ZZ1 |_NULL <_FO |_NULL Hi_UH |_NULL ._. 
</s>
<s>
In_II a_AT1 similar_JJ fashion_NN1 ,_, the_AT update_NN1 gate_NN1 utl_NN1 learns_VVZ to_TO decide_VVI what_DDQ part_NN1 of_IO ht_NNU 1l_NNU will_VM be_VBI leaked_VVN into_II the_AT computation_NN1 of_IO the_AT current_JJ hidden_JJ state_NN1 htl_NNU ._. 
</s>
<s>
Process_VV0 tools_NN2 that_CST capture_VV0 practitioners_NN2 '_NULL communications_NN2 (_( such_II21 as_II22 IBM_NP1 Rational_JJ Jazz_NN1 )_) may_VM also_RR benefit_VVI from_II attitude_NN1 and_CC behavior_NN1 visualizations_NN2 ._. 
</s>
<s>
For_IF simplicity_NN1 of_IO presentation_NN1 we_PPIS2 assume_VV0 fairness_NN1 is_VBZ provided_VVN above_II the_AT MAC_NN1 layer_NN1 in_II the_AT current_JJ paper_NN1 ,_, excluding_VVG the_AT MAC_NN1 and_CC PHY_JJ overhead_NN1 ._. 
</s>
<s>
Implementation_NN1 can_VM not_XX be_VBI viewed_VVN as_II a_AT1 static_JJ execution_NN1 of_IO plans_NN2 ,_, and_CC like_II other_JJ managed_JJ activities_NN2 require_VV0 performance_NN1 metrics_NN2 ,_, periodic_JJ revision_NN1 of_IO plans_NN2 ,_, and_CC incentives_NN2 for_IF achieving_VVG security_NN1 goals_NN2 ._. 
</s>
<s>
JLev_NP1 and_CC JSed_NP1 are_VBR weakly_RR correlated_VVN and_CC have_VH0 no_AT relevant_JJ positive_JJ correlations_NN2 with_IW the_AT other_JJ objectives_NN2 ._. 
</s>
<s>
The_AT overall_JJ query_NN1 selectivity_NN1 can_VM not_XX be_VBI used_VVN to_TO estimate_VVI the_AT data_NN transfer_NN1 ._. 
</s>
<s>
The_AT categorisation_NN1 of_IO both_DB2 Evie_NP1 and_CC ,_, in_II turn_NN1 ,_, Seb_NP1 as_CSA having_VHG attachment_NN1 difficulties_NN2 ,_, for_REX21 instance_REX22 ,_, and_CC of_IO Seb_NP1 being_VBG "_" at_II risk_NN1 "_" ._. 
</s>
<s>
The_AT authenticated_JJ delegation_NN1 lists_NN2 and_CC authentication_NN1 delegation_NN1 trees_NN2 are_VBR more_DAR efficientboth_NN1 require_VV0 at_II times_NNT2 an_AT1 order_NN1 of_IO magnitude_NN1 less_DAR computation_NN1 than_CSN simple_JJ attestations_NN2 ._. 
</s>
<s>
Teams_NN2 may_VM not_XX be_VBI able_JK to_TO achieve_VVI the_AT required_JJ level_NN1 of_IO interdependent_JJ working_JJ or_CC have_VH0 the_AT qualities_NN2 that_CST have_VH0 been_VBN found_VVN to_TO characterize_VVI high-performing_JJ teams_NN2 ._. 
</s>
<s>
Especially_RR in_II the_AT case_NN1 of_IO administrative_JJ deliveries_NN2 ,_, legal_JJ issues_NN2 will_VM have_VHI to_TO be_VBI resolved_VVN by_II introducing_VVG common_JJ rules_NN2 and_CC legislations_NN2 for_IF cross-border_JJ certified_JJ mail_NN1 ._. 
</s>
<s>
In_RR21 addition_RR22 ,_, because_CS this_DD1 technique_NN1 does_VDZ not_XX approximate_VVI responses_NN2 but_CCB rather_RR classify_VV0 them_PPHO2 as_CSA failed_VVN or_CC safe_JJ ,_, it_PPH1 naturally_RR handles_VVZ discontinuities_NN2 ._. 
</s>
<s>
When_CS retrieving_VVG service_NN1 descriptions_NN2 ,_, the_AT ontologies_NN2 their_APPGE annotations_NN2 refer_VV0 to_II are_VBR also_RR retrieved_VVN so_RR that_DD1 reasoning_NN1 can_VM be_VBI performed_VVN ._. 
</s>
<s>
Control_NN1 is_VBZ somewhat_RR different_JJ from_II manipulation_NN1 since_CS in_II control_NN1 problems_NN2 we_PPIS2 change_VV0 the_AT structure_NN1 of_IO the_AT election_NN1 (_( number_NN1 of_IO candidates_NN2 ,_, number_NN1 of_IO voters_NN2 ,_, etc_RA ._. )_) ._. 
</s>
<s>
If_CS we_PPIS2 consider_VV0 the_AT added_JJ efficiency_NN1 because_II21 of_II22 the_AT use_NN1 of_IO IS_VBZ when_RRQ the_AT combined_JJ framework_NN1 is_VBZ chosen_VVN ,_, then_RT the_AT computational_JJ advantages_NN2 from_II using_VVG this_DD1 framework_NN1 are_VBR even_RR higher_JJR :_: in_II this_DD1 example_NN1 the_AT computational_JJ cost_NN1 of_IO each_DD1 iteration_NN1 of_IO SPSA_NN1 is_VBZ 8_MC and_CC 11_MC times_NNT2 smaller_JJR for_IF problems_NN2 D1_FO and_CC D2_FO ,_, respectively_RR ,_, in_II the_AT setting_NN1 of_IO the_AT combined_JJ framework_NN1 ._. 
</s>
<s>
For_IF each_DD1 state_NN1 evaluated_VVN ,_, FF-Hindsight_NN1 creates_VVZ sets_NN2 of_IO time-varying_JJ classical_JJ planning_NN1 problems_NN2 and_CC uses_VVZ FF_NNU to_TO find_VVI which_DDQ portions_NN2 are_VBR solvable_JJ ._. 
</s>
<s>
To_II this_DD1 effect_NN1 we_PPIS2 want_VV0 to_TO avoid_VVI poorly_RR proportioned_VVN elements_NN2 ,_, e.g._REX thin_JJ triangles_NN2 (_( Fig._NN1 3_MC ,_, top_NN1 )_) or_CC thin_JJ polygons_NN2 (_( Fig._NN1 3_MC ,_, bottom_NN1 )_) ._. 
</s>
<s>
The_AT costs_NN2 ,_, both_RR in_II31 terms_II32 of_II33 repairs_NN2 and_CC lost_JJ productivity_NN1 ,_, were_VBDR on_II the_AT order_NN1 of_IO 10_MC billion_NNO dollars_NNU2 &lsqb;_( 72_MC &rsqb;_) ._. 
</s>
<s>
The_AT surveys_NN2 themselves_PPX2 are_VBR implemented_VVN as_CSA XML_NP1 documents_NN2 ._. 
</s>
<s>
So_RG far_RR ,_, with_IW this_DD1 approach_NN1 ,_, only_RR less_DAR than_CSN one_MC1 millimeter_JJR distances_NN2 of_IO laser_NN1 propagation_NN1 were_VBDR successfully_RR computed_VVN ._. 
</s>
<s>
The_AT first_MD question_NN1 is_VBZ how_RRQ to_TO initialise_VVI K_ZZ1 and_CC P._NP1 We_PPIS2 have_VH0 developed_VVN an_AT1 initialisation_NN1 approach_NN1 that_CST exploits_NN2 information_NN1 from_II reasoning_VVG tests_NN2 in_II order_NN1 to_TO eagerly_RR identify_VVI subsumption_NN1 relations_NN2 and_CC unsatisfiable_JJ classes_NN2 and_CC thus_RR reduce_VV0 the_AT overall_JJ amount_NN1 of_IO work_NN1 ._. 
</s>
<s>
Reactive_JJ online_JJ management_NN1 is_VBZ supported_VVN through_II local_JJ processing_NN1 and_CC decision_NN1 making_VVG within_II the_AT sensor_NN1 network_NN1 ._. 
</s>
<s>
This_DD1 model_NN1 recognizes_VVZ the_AT seemingly_RR irrational_JJ behavior_NN1 that_CST is_VBZ all_DB too_RG common_JJ (_( White_NP1 ,_, 2013_MC )_) ._. 
</s>
<s>
Continued_JJ improvements_NN2 in_II best_JJT practices_NN2 to_II *_FU increase_NN1 security_NN1 costs_VVZ by_II an_AT1 amount_NN1 (_( c(*)@@@c(1)_FO )_) which_DDQ is_VBZ equal_JJ to_II the_AT decrease_NN1 in_II expected_JJ losses_NN2 ._. 
</s>
<s>
In_II the_AT case_NN1 of_IO our_APPGE embedded_JJ system_NN1 ,_, the_AT sender_NN1 and_CC the_AT receiver_NN1 are_VBR cores_NN2 ,_, and_CC the_AT shared_JJ resource_NN1 is_VBZ the_AT state_NN1 of_IO the_AT policy_NN1 ._. 
</s>
<s>
The_AT arguments_NN2 concerning_II the_AT routing_NN1 properties_NN2 that_CST were_VBDR used_VVN in_II Section_NN1 4_MC ,_, to_TO explain_VVI why_RRQ the_AT QoS_NN2 in_II delivering_VVG packets_NN2 to_II their_APPGE destinations_NN2 depends_VVZ on_II the_AT ecf_NN1 type_NN1 and_CC it_PPH1 changes_VVZ with_IW the_AT source_NN1 load_NN1 values_NN2 ,_, are_VBR applicable_JJ also_RR here_RL to_TO explain_VVI the_AT observed_JJ differences_NN2 among_II the_AT dynamic_JJ QoSs_NN1 ._. 
</s>
<s>
Their_APPGE filter_NN1 evaluation_NN1 toolkit_NN1 ,_, given_VVN a_AT1 corpus_NN1 and_CC a_AT1 filter_NN1 ,_, compares_VVZ the_AT filter_NN1 classification_NN1 of_IO each_DD1 message_NN1 with_IW the_AT gold_NN1 standard_NN1 to_TO report_VVI effectiveness_NN1 measures_NN2 with_IW 95%_NNU confidence_NN1 limits_NN2 ._. 
</s>
<s>
It_PPH1 is_VBZ important_JJ that_CST we_PPIS2 evaluate_VV0 the_AT high-speed_JJ TCPs_NP1 in_II various_JJ other_JJ different_JJ environment_NN1 ._. 
</s>
<s>
Then_RT ,_, according_II21 to_II22 the_AT proposed_JJ relaxed_JJ shape_NN1 constraint_NN1 and_CC the_AT new_JJ stopping_VVG function_NN1 described_VVN in_II Section_NN1 3.2_MC ,_, we_PPIS2 obtain_VV0 the_AT sketch_NN1 map_NN1 of_IO the_AT relaxed_JJ shape_NN1 constraint_NN1 in_II Fig._NN1 6(e)_FO ._. 
</s>
<s>
The_AT conceptual_JJ model_NN1 and_CC vocabulary_NN1 deliverables_NN2 were_VBDR set_VVN to_TO become_VVI W3C_FO Recommendations_NN2 ,_, for_IF which_DDQ there_EX is_VBZ a_AT1 burden_NN1 of_IO proof_NN1 of_IO implementability_NN1 and_CC inter-operability_NN1 ,_, whereas_CS the_AT other_JJ documents_NN2 became_VVD W3C_FO Notes_NN2 ,_, technical_JJ documents_NN2 without_IW such_DA a_AT1 requirement_NN1 but_CCB still_RR approved_VVN by_II Working_JJ Group_NN1 consensus_NN1 ._. 
</s>
<s>
We_PPIS2 provide_VV0 an_AT1 algorithm_NN1 to_TO do_VDI this_DD1 ._. 
</s>
<s>
In_II the_AT Round_JJ Robin_NN1 scheme_NN1 ,_, the_AT state_NN1 of_IO the_AT AP_NP1 strictly_RR cycles_VVZ through_II the_AT states_NN2 in_II round_JJ robin_NN1 order_NN1 ,_, i.e._REX 1,2_MC ,_, ,_, L_ZZ1 ,_, 1,2_MC ,_, ,_, L_ZZ1 ,_, 1,2._MC ._. 
</s>
<s>
In_II a_AT1 practical_JJ sense_NN1 ,_, information_NN1 security_NN1 awareness_NN1 programs_NN2 can_VM include_VVI the_AT use_NN1 of_IO information_NN1 flyers/leaflets_NN2 ,_, brochures_NN2 and_CC websites_NN2 ;_; school_NN1 seminars_NN2 and_CC talks_NN2 ;_; community_NN1 workshops_NN2 and_CC focus_VV0 groups_NN2 ;_; e-security_JJ merchandise_NN1 ;_; financial_JJ support_NN1 grants_NN2 ;_; and_CC ,_, public_JJ competition_NN1 and_CC quiz_NN1 contests_NN2 (_( Thomson_NP1 and_CC von_NP1 Solms_NP1 ,_, 1998_MC and_CC Wooding_NP1 et_RA21 al._RA22 ,_, 2003_MC )_) ._. 
</s>
<s>
Therefore_RR ,_, given_VVN some_DD value_NN1 C_ZZ1 ,_, we_PPIS2 can_VM calculate_VVI the_AT equation_NN1 above_RL to_TO determine_VVI the_AT boundary_NN1 ._. 
</s>
<s>
In_BCL21 order_BCL22 to_TO perform_VVI a_AT1 benchmark_NN1 comparison_NN1 of_IO SW_ND1 ,_, we_PPIS2 also_RR provide_VV0 a_AT1 discussion_NN1 of_IO DTW_NP1 (_( Section_NN1 5_MC )_) and_CC the_AT discrete_JJ HMM_UH (_( Section_NN1 6_MC )_) and_CC factors_NN2 involved_JJ with_IW their_APPGE application_NN1 in_II spatial_JJ activity_NN1 recognition_NN1 ._. 
</s>
<s>
Backpropagation_NN1 updates_VVZ the_AT records_NN2 associated_VVN with_IW the_AT actions_NN2 played_VVN during_II the_AT playout(MAST-6)_NN1 ,_, with_IW the_AT player_NN1 who_PNQS played_VVD each_DD1 action_NN1 as_CSA contextual_JJ information(MAST-5)_NN1 ._. 
</s>
<s>
If_CS we_PPIS2 accept_VV0 the_AT argument_NN1 that_CST modern_JJ ,_, economically_RR developed_JJ societies_NN2 are_VBR increasingly_RR becoming_VVG '_NULL information_NN1 societies_NN2 '_NULL ,_, then_RT ,_, it_PPH1 follows_VVZ that_CST threats_NN2 to_II information_NN1 can_VM be_VBI seen_VVN as_CSA threats_NN2 to_II the_AT core_NN1 of_IO these_DD2 societies_NN2 (_( Eriksson_NP1 and_CC Giacomello_NP1 ,_, 2006_MC )_) ._. 
</s>
<s>
Initially_RR ,_, FreightX_NP1 '_NULL s_ZZ1 executives_NN2 did_VDD not_XX charge_VVI membership_NN1 fees_NN2 in_II exchange_NN1 for_IF website_NN1 access_NN1 ._. 
</s>
<s>
Consequently_RR ,_, NDP2_FO '_NULL s_ZZ1 performance_NN1 is_VBZ in_II some_DD ways_NN2 better_RRR than_CSN MBP_NP1 '_NULL s_ZZ1 (_( e.g._REX ,_, how_RGQ many_DA2 problems_NN2 it_PPH1 could_VM solve_VVI )_) ,_, and_CC in_II some_DD ways_NN2 worse_JJR than_CSN MBP_NP1 '_NULL s_ZZ1 (_( e.g._REX ,_, the_AT amount_NN1 of_IO CPU_NN1 time_NNT1 it_PPH1 used_JJ )_) ._. 
</s>
<s>
JIRA_NP1 provides_VVZ the_AT following_JJ easily_RR accessible_JJ issue_NN1 information_NN1 (_( properties_NN2 )_) ,_, namely:Project_FO Name_NN1 ,_, Title_NN1 ,_, Description_NN1 ,_, Summary_NN1 ,_, Type_NN1 ,_, Priority_NN1 ,_, State_VV0 ,_, Resolution_NN1 ,_, Reporter_NN1 ,_, Assignee_NP1 ._. 
</s>
<s>
Besides_RR developing_VVG new_JJ algorithms_NN2 for_IF the_AT functionality_NN1 already_RR included_VVN (_( e.g._REX ,_, for_IF linking_VVG and_CC feature_NN1 selection_NN1 )_) ,_, there_EX are_VBR also_RR some_DD new_JJ functionalities_NN2 currently_RR being_VBG investigated_VVN ._. 
</s>
<s>
We_PPIS2 also_RR restate_VV0 here_RL a_AT1 simple_JJ ,_, useful_JJ lemma_NN1 of_IO the_AT Gao-Rexford_NP1 conditions_NN2 proved_VVN by_II Gao_NP1 ,_, Griffin_NP1 and_CC Rexford_NP1 in_II &lsqb;_( 16_MC &rsqb;_) ._. 
</s>
<s>
Annotations_NN2 are_VBR provided_VVN in_II a_AT1 proprietary_JJ XML_NP1 format_NN1 shown_VVN in_II Fig._NN1 1_MC1 ._. 
</s>
<s>
They_PPHS2 are_VBR snapping_VVG your_APPGE text_NN1 to_II the_AT nearest_JJT word_NN1 in_II the_AT "_" grid_NN1 "_" provided_VVN by_II the_AT dictionary_NN1 ._. 
</s>
<s>
When_CS a_AT1 DPS_NP1 is_VBZ consumed_VVN ,_, values_NN2 contained_VVN within_II the_AT fields_NN2 of_IO an_AT1 instance_NN1 of_IO a_AT1 message_NN1 may_VM be_VBI substituted_VVN with_IW their_APPGE referenced_JJ variables_NN2 at_II both_DB2 the_AT semantic_JJ and_CC data_NN acquisition_NN1 layers_NN2 ._. 
</s>
<s>
In_II the_AT next_MD section_NN1 we_PPIS2 also_RR compare_VV0 our_APPGE method_NN1 with_IW related_JJ approaches_NN2 like_II topology_NN1 adaptive_JJ snake_NN1 &lsqb;_( 13_MC &rsqb;_) ,_, active_JJ net_NN1 &lsqb;_( 2_MC &rsqb;_) and_CC level_JJ set_NN1 &lsqb;_( 3_MC &rsqb;_) ,_, &lsqb;_( 4_MC &rsqb;_) ,_, &lsqb;_( 9_MC &rsqb;_) ._. 
</s>
<s>
In_RR21 addition_RR22 ,_, many_DA2 different_JJ implantable_JJ sensor_NN1 arrays_NN2 have_VH0 been_VBN reported_VVN ,_, which_DDQ could_VM provide_VVI more_RGR spatial_JJ information_NN1 &lsqb;_( 12_MC &rsqb;_) ,_, &lsqb;_( 13_MC &rsqb;_) ,_, &lsqb;_( 14_MC &rsqb;_) and_CC &lsqb;_( 15_MC &rsqb;_) ._. 
</s>
<s>
The_AT training_NN1 set_NN1 is_VBZ then_RT used_VVN to_TO search_VVI for_IF the_AT nearest_JJT neighbors_NN2 ._. 
</s>
<s>
The_AT current_JJ version_NN1 of_IO the_AT model_NN1 does_VDZ not_XX implement_VVI this_DD1 concern_NN1 ,_, while_CS achieving_VVG reliable_JJ projection_NN1 accuracy_NN1 ._. 
</s>
<s>
In_II 8-ball_NN1 ,_, since_CS a_AT1 player_NN1 may_VM aim_VVI at_II any_DD of_IO his_APPGE assigned_JJ solids_NN2 or_CC stripes_NN2 ,_, there_EX are_VBR usually_RR straight-in_JJ shots_NN2 available_JJ ._. 
</s>
<s>
Adding_VVG this_DD1 constraint_NN1 resulted_VVN in_II just_RR two_MC candidate_NN1 models_NN2 ,_, namely_REX one_PN1 using_VVG the_AT Hamon_NP1 PET_NN1 equation_NN1 and_CC the_AT other_NN1 using_VVG the_AT Hargreaves_NP1 PET_NN1 equation_NN1 in_II all_DB seven_MC sub-compartments_NN2 ._. 
</s>
<s>
One_MC1 type_NN1 of_IO chains_NN2 in_II the_AT semantic_JJ drift_NN1 domain_NN1 is_VBZ morphing_VVG chains_NN2 ,_, a_AT1 term_NN1 coined_VVD in_II &lsqb;_( 7_MC &rsqb;_) ._. 
</s>
<s>
TSSTM&lsqb;7&rsqb;was_FO our_APPGE base_NN1 stress_NN1 test_NN1 methodology_NN1 ,_, which_DDQ requires_VVZ the_AT timing_NN1 information_NN1 of_IO the_AT messages_NN2 in_II the_AT system_NN1 to_TO be_VBI deterministic_JJ ._. 
</s>
<s>
Indeed_RR ,_, in_II the_AT former_DA cases_NN2 ,_, the_AT training_NN1 data_NN and_CC the_AT test_NN1 data_NN come_VV0 exactly_RR from_II the_AT same_DA uniform_JJ distribution_NN1 ,_, the_AT one_PN1 across_II the_AT language_NN1 that_CST can_VM be_VBI generated_VVN by_II our_APPGE grammar_NN1 ,_, while_CS the_AT 4.19%_FO of_IO Average_JJ Per-Formula_NP1 Accuracy_NN1 registered_VVN by_II the_AT Grammar_NN1 Parser_NN1 baseline_NN1 indicates_VVZ that_CST the_AT difference_NN1 between_II the_AT bootstrap_NN1 training_NN1 data_NN and_CC the_AT test_NN1 data_NN from_II the_AT 425M_NNU reference_NN1 set_NN1 is_VBZ remarkable_JJ ._. 
</s>
<s>
For_IF smaller_JJR radii_NN2 ,_, only_RR the_AT first_MD few_DA2 regular_JJ polygons_NN2 can_VM actually_RR be_VBI reliably_RR discriminated_VVN from_II the_AT image_NN1 information_NN1 due_II21 to_II22 pixel_NN1 quantisation_NN1 ._. 
</s>
<s>
We_PPIS2 define_VV0 grid_NN1 productivity_NN1 as_II the_AT uninterrupted_JJ continuity_NN1 of_IO any_DD scientific_JJ activity_NN1 over_II the_AT grid_NN1 ._. 
</s>
<s>
The_AT query_NN1 reply_NN1 cost_NN1 depends_VVZ on_II the_AT reply_NN1 scheme_NN1 ._. 
</s>
<s>
One_MC1 questionnaire_NN1 was_VBDZ based_VVN on_II the_AT benefits_NN2 and_CC problems_NN2 most_RRT associated_VVN with_IW following_VVG a_AT1 mapping_NN1 study_NN1 with_IW another_DD1 mapping_NN1 study_NN1 of_IO broader_JJR scope_NN1 ._. 
</s>
<s>
However_RR ,_, the_AT quality_NN1 score_NN1 for_IF studies_NN2 published_VVN in_II recent_JJ years_NNT2 is_VBZ consistently_RR high_RR indicating_VVG that_CST the_AT researchers_NN2 are_VBR providing_VVG more_DAR details_NN2 in_II31 terms_II32 of_II33 study_NN1 design_NN1 and_CC execution_NN1 ._. 
</s>
<s>
The_AT transformation_NN1 A_ZZ1 can_VM be_VBI used_VVN to_TO map_VVI the_AT points_NN2 of_IO a_AT1 rule_NN1 over_II the_AT standard_JJ triangle_NN1 to_II R._NP1 The_AT MIP_NN1 formulation_NN1 (_( 5–12_MCMC )_) was_VBDZ implemented_VVN using_VVG open_JJ source_NN1 lpsolve1_FO 5.5_MC dynamic_JJ link_NN1 library_NN1 (_( DLL_NP1 )_) and_CC C++_FO programming_NN1 language_NN1 ._. 
</s>
<s>
Equation_NN1 (_( 3_MC )_) is_VBZ based_VVN on_II the_AT empirically_RR based_VVN assumption_NN1 that_CST variation_NN1 of_IO the_AT river_NN1 level_NN1 affects_VVZ fish_NN population_NN1 dynamics_NN (_( Welcomme_NP1 ,_, 1985_MC ,_, Merona_NP1 and_CC Gascuel_NP1 ,_, 1993_MC ,_, Batista_NP1 and_CC Petrere_NP1 ,_, 2007_MC )_) as_RG much_DA1 as_CSA changes_NN2 in_II fishing_NN1 pressure_NN1 ._. 
</s>
<s>
For_REX21 instance_REX22 ,_, while_CS the_AT solid_JJ blue_JJ node_NN1 in_II Fig._NN1 3(a)_FO is_VBZ likely_JJ to_TO play_VVI a_AT1 significant_JJ role_NN1 in_II spreading_NN1 (_( duplicating_VVG )_) knowledge_NN1 across_II this_DD1 local_JJ network_NN1 segment_NN1 ,_, this_DD1 knowledge_NN1 would_VM only_RR remain_VVI local_JJ in_II the_AT absence_NN1 of_IO the_AT other_JJ members_NN2 '_NULL connections_NN2 to_II Fig._NN1 3(b)_FO the_AT green_JJ node_NN1 ._. 
</s>
<s>
If_CS an_AT1 antibody_NN1 '_NULL s_ZZ1 survival_NN1 time_NNT1 exceeds_VVZ the_AT SurviveTime_NNT1 ,_, it_PPH1 will_VM be_VBI replaced_VVN by_II a_AT1 new_JJ randomly_RR generated_VVN antibody_NN1 ._. 
</s>
<s>
The_AT faces_NN2 (_( or_CC edges_NN2 )_) of_IO these_DD2 meshes_NN2 are_VBR spatially_RR conforming_VVG within_II each_DD1 subdomain_NN1 ,_, but_CCB are_VBR allowed_VVN to_TO be_VBI non-conforming_JJ along_II subdomain_NN1 interfaces_NN2 ._. 
</s>
<s>
Consider_VV0 then_RT a_AT1 configuration_NN1 @h∈Vhd_FO for_IF which_DDQ the_AT exact_JJ bilinear_JJ form_NN1 BB_NP1 is_VBZ coercive_JJ ._. 
</s>
<s>
We_PPIS2 considered_VVD a_AT1 set_NN1 of_IO data_NN consisting_VVG of_IO multi-oriented_JJ and_CC multi-scale_JJ text_NN1 characters_NN2 to_TO perform_VVI this_DD1 character_NN1 recognition_NN1 evaluation_NN1 ._. 
</s>
<s>
We_PPIS2 also_RR classify_VV0 CMS_NP1 according_II21 to_II22 this_DD1 criterion_NN1 because_CS different_JJ approaches_NN2 may_VM have_VHI different_JJ demands_NN2 in_II31 terms_II32 of_II33 infrastructural_JJ requirements_NN2 and_CC are_VBR thus_RR of_IO interest_NN1 when_CS evaluating_VVG systems_NN2 toward_II interoperability_NN1 ._. 
</s>
<s>
Similarly_RR ,_, user_NN1 interface_NN1 design_NN1 plays_VVZ a_AT1 significant_JJ role_NN1 in_II avoiding_VVG errors_NN2 involving_VVG the_AT use_NN1 of_IO technology_NN1 ._. 
</s>
<s>
Such_DA computer_NN1 devices_NN2 in_II the_AT human_JJ interaction_NN1 loop_NN1 would_VM possibly_RR provide_VVI feedback_NN1 to_TO enhance_VVI the_AT human–human_NN1 communication_NN1 and_CC relations_NN2 in_RR21 general_RR22 &lsqb;_( 1_MC1 &rsqb;_) ,_, &lsqb;_( 2_MC &rsqb;_) ,_, &lsqb;_( 3_MC &rsqb;_) ,_, &lsqb;_( 4_MC &rsqb;_) ,_, &lsqb;_( 5_MC &rsqb;_) ,_, &lsqb;_( 6_MC &rsqb;_) ,_, &lsqb;_( 7_MC &rsqb;_) ,_, concurring_VVG to_II the_AT development_NN1 of_IO the_AT so-called_JJ human_JJ computing_NN1 scientific_JJ area_NN1 &lsqb;_( 8_MC &rsqb;_) ._. 
</s>
<s>
Unlike_II the_AT method_NN1 introduced_VVN in_II &lsqb;_( 1_MC1 &rsqb;_) and_CC those_DD2 reviewed_VVN in_II &lsqb;_( 2_MC &rsqb;_) ,_, we_PPIS2 do_VD0 not_XX establish_VVI the_AT graph_NN1 representations_NN2 based_VVN on_II the_AT original_JJ vertex_NN1 set_NN1 of_IO a_AT1 hypergraph_NN1 ._. 
</s>
<s>
By_II using_VVG sleeper_NN1 malware_NN1 that_CST hides_VVZ in_II a_AT1 system_NN1 and_CC activities_NN2 upon_II receiving_VVG a_AT1 signal_NN1 ,_, the_AT principle_NN1 of_IO surprise_NN1 can_VM be_VBI applied_VVN ._. 
</s>
<s>
Of_RR21 course_RR22 ,_, this_DD1 means_VVZ that_CST he_PPHS1 has_VHZ to_TO restrict_VVI his_APPGE attention_NN1 to_II basic_JJ formulas_NN2 ._. 
</s>
<s>
The_AT difference_NN1 lies_VVZ in_II the_AT selection_NN1 of_IO theconn-slotat_NN1 a_AT1 particular_JJ position_NN1 for_IF reservation_NN1 at_II each_DD1 link_NN1 ._. 
</s>
<s>
Mau_NP1 and_CC Winter_NP1 (_( 1997_MC )_) found_VVD a_AT1 value_NN1 of_IO 0.85_MC to_TO be_VBI most_RGT appropriate_JJ and_CC Tan_JJ et_NP1 ?_? al_NP1 ._. 
</s>
<s>
(_( 2009a_FO )_) suggested_VVD using_VVG the_AT recession_NN1 constant_NN1 as_II the_AT filter_NN1 parameter_NN1 value_NN1 ,_, which_DDQ varies_VVZ from_II catchment_NN1 to_II catchment_NN1 ._. 
</s>
<s>
Finally_RR ,_, this_DD1 work_NN1 is_VBZ concluded_VVN and_CC future_JJ works_NN are_VBR discussed_VVN in_II Section_NN1 7_MC ._. 
</s>
<s>
In_II31 keeping_II32 with_II33 the_AT aim_NN1 of_IO this_DD1 special_JJ issue_NN1 ,_, the_AT field_NN1 study_NN1 aspect_NN1 of_IO this_DD1 research_NN1 is_VBZ noteworthy_JJ ._. 
</s>
<s>
Model_NN1 formulation_NN1 enables_VVZ the_AT optimizer_NN1 to_TO determine_VVI destination_NN1 of_IO mined_JJ parcels_NN2 rather_II21 than_II22 using_VVG pre-defined_JJ destinations_NN2 ._. 
</s>
<s>
An_AT1 instantiation_NN1 of_IO a_AT1 Lock_NN1 Type_NN1 Graph_NN1 is_VBZ called_VVN a_AT1 Lock_NN1 Instance_NN1 Graph_NN1 ._. 
</s>
<s>
Compliance_NN1 with_IW standards_NN2 also_RR frees_VVZ companies_NN2 from_II the_AT constraints_NN2 of_IO proprietary_JJ formats_NN2 when_CS choosing_VVG knowledge_NN1 management_NN1 software_NN1 ._. 
</s>
<s>
We_PPIS2 also_RR observed_VVD that_CST participants_NN2 used_VVD the_AT available_JJ rating_NN1 and_CC ranking_VVG scales_NN2 inconsistently_RR ._. 
</s>
<s>
On_II its_APPGE own_DA it_PPH1 does_VDZ not_XX explain_VVI very_RG much_DA1 ,_, but_CCB provides_VVZ a_AT1 method_NN1 to_TO help_VVI explain_VVI why_RRQ things_NN2 work_VV0 (_( or_CC why_RRQ they_PPHS2 do_VD0 not_XX work_VVI )_) "_" The_AT aim_NN1 of_IO this_DD1 paper_NN1 is_VBZ to_TO achieve_VVI user_NN1 privacy_NN1 without_IW requiring_VVG a_AT1 trusted_JJ platform_NN1 or_CC involving_VVG a_AT1 trusted_JJ third_MD party_NN1 ._. 
</s>
<s>
Learner_NN1 performance_NN1 may_VM be_VBI tracked_VVN by_II serious_JJ games_NN2 and_CC other_JJ adaptive_JJ technologies_NN2 ._. 
</s>
<s>
Thefeedback_NN1 widthofPis_NN1 the_AT size_NN1 of_IO a_AT1 smallest_JJT setV_NN1 of_IO atoms_NN2 such_CS21 that_CS22 every_AT1 cycle_NN1 ofU(P)runs_VVZ through_II an_AT1 atom_NN1 inV_NN1 (_( such_DA a_AT1 setVis_NN1 called_VVN afeedback_NN1 vertex_NN1 set_NN1 )_) ._. 
</s>
<s>
The_AT method_NN1 was_VBDZ improved_VVN by_II adding_VVG an_AT1 artificial_JJ stress_NN1 term_NN1 for_IF reducing_VVG the_AT so-called_JJ tensile_JJ instability_NN1 effect_NN1 ._. 
</s>
<s>
They_PPHS2 observed_VVD that_CST the_AT codes_NN2 with_IW high_JJ level_NN1 of_IO Lsym_NP1 are_VBR more_RGR discriminative_JJ than_CSN those_DD2 with_IW low_JJ level_NN1 of_IO Lsym_NN1 by_II qualitative_JJ visual_JJ inspection_NN1 ._. 
</s>
<s>
Initially_RR a_AT1 small_JJ square_JJ matrix_NN1 (_( i.e._REX ,_, n_ZZ1 ?_? n_ZZ1 mask_NN1 )_) is_VBZ slid_VVN over_II the_AT original_JJ gray_JJ level_JJ image_NN1 in_II x_ZZ1 direction_NN1 and_CC then_RT in_II y_ZZ1 direction_NN1 ,_, with_IW step_NN1 of_IO one_MC1 pixel_NN1 ._. 
</s>
<s>
For_REX21 example_REX22 ,_, in_II our_APPGE experiments_NN2 (_( iteration_NN1 1_MC1 in_II Section5.3_FO )_) ,_, we_PPIS2 used_VVD the_AT IBM_NP1 Rational_JJ Software_NN1 Architect_NN1 (_( RSA_NP1 )_) which_DDQ was_VBDZ able_JK to_TO successfully_RR achieve_VVI 100%_NNU automated_JJ synchronization_NN1 (_( i.e._REX ,_, there_EX was_VBDZ no_AT need_NN1 for_IF manual_JJ changes_NN2 )_) ._. 
</s>
<s>
One_MC1 area_NN1 where_CS we_PPIS2 had_VHD difficulty_NN1 was_VBDZ the_AT MPI.Filewrite()_NP1 ._. 
</s>
<s>
A_AT1 point_NN1 to_II much_RR the_AT same_DA effect_NN1 was_VBDZ made_VVN by_II Lindenschmidt_NP1 (_( 2001_MC )_) but_CCB the_AT rather_RG famous_JJ Einstein_NP1 quotation_NN1 ,_, "_" As_CS31 far_CS32 as_CS33 the_AT laws_NN2 of_IO mathematics_NN1 refer_VV0 to_II reality_NN1 ,_, they_PPHS2 are_VBR not_XX certain_JJ ;_; and_CC as_CS31 far_CS32 as_CS33 they_PPHS2 are_VBR certain_JJ ,_, they_PPHS2 do_VD0 not_XX refer_VVI to_II reality_NN1 ._. "_" 
</s>
<s>
The_AT convergence_NN1 rate_NN1 of_IO the_AT scheme_NN1 for_IF larger_JJR mesh_NN1 sizes_NN2 O(δ2.43)_FO is_VBZ dominated_VVN by_II the_AT largest_JJT neglected_JJ term_NN1 in_II the_AT tip_NN1 asymptotic_JJ expansion_NN1 ._. 
</s>
<s>
It_PPH1 was_VBDZ found_VVN that_CST violations_NN2 of_IO the_AT low-coverage_JJ regime_NN1 were_VBDR due_JJ exclusively_RR to_II high_JJ CO_NP1 (_( carbon_NN1 monoxide_NN1 )_) coverage_NN1 at_II relatively_RR low_JJ temperatures_NN2 and_CC high_JJ C_NP1 coverage_NN1 at_II relatively_RR high_JJ temperatures_NN2 ,_, as_CSA shown_VVN in_II Fig._NN1 3_MC ._. 
</s>
<s>
This_DD1 is_VBZ likely_JJ due_II21 to_II22 the_AT extreme_JJ lengths_NN2 of_IO the_AT adversary_NN1 '_NULL s_ZZ1 paths_NN2 ._. 
</s>
<s>
The_AT graphs_NN2 used_VVN in_II the_AT remaining_JJ sections_NN2 were_VBDR generated_VVN in_II an_AT1 analogous_JJ way_NN1 ,_, save_CS21 that_CS22 the_AT source_NN1 data_NN may_VM have_VHI been_VBN taken_VVN from_II different_JJ dates_NN2 ._. 
</s>
<s>
Mode_NN1 of_IO node_NN1 degree_NN1 distribution_NN1 is_VBZ not_XX helpful_JJ however_RR in_II detecting_VVG skeletons_NN2 supporting_VVG localized_JJ oscillators_NN2 ._. 
</s>
<s>
N_ZZ1 is_VBZ the_AT number_NN1 of_IO points_NN2 ,_, nopt_NN1 and_CC ISEopt_NP1 are_VBR the_AT values_NN2 obtained_VVN for_IF the_AT reference_NN1 approximation_NN1 ,_, and_CC ndp_NNU and_CC ISE_NP1 are_VBR those_DD2 obtained_VVN for_IF the_AT method_NN1 being_VBG evaluated_VVN ._. 
</s>
<s>
For_IF each_DD1 of_IO these_DD2 tags_NN2 ,_, the_AT server_NN1 may_VM specify_VVI anamewhich_NN1 names_VVZ the_AT parameter_NN1 being_VBG collected_VVN from_II the_AT user_NN1 and_CC a_AT1 defaultvalue_NN1 ._. 
</s>
<s>
Firstly_RR ,_, an_AT1 Euclidean_JJ matching_JJ function_NN1 d_ZZ1 (_( ai_NNU ,_, bj_NNU )_) and_CC corresponding_JJ matching_JJ threshold_NN1 θ_NULL were_VBDR proposed_VVN ._. 
</s>
<s>
As_II a_AT1 result_NN1 ,_, it_PPH1 took_VVD a_AT1 long_JJ time_NNT1 to_TO fix_VVI the_AT bug_NN1 ._. 
</s>
<s>
These_DD2 additional_JJ search_NN1 strings_NN2 resulted_VVN in_II the_AT identification_NN1 of_IO a_AT1 total_NN1 of_IO 718_MC papers_NN2 ._. 
</s>
<s>
Let_VV0 v_ZZ1 be_VBI a_AT1 satisfying_JJ assignment_NN1 for_IF f_ZZ1 and_CC let_VV0 (_( v_ZZ1 )_) be_VBI a_AT1 valuation_NN1 for_IF B_ZZ1 on_II I_ZZ1 where_RRQ y1=a_FO ,_, y2=b_FO and_CC if_CS v_ZZ1 mapsxi_NN2 to_II 1_MC1 then_JJ wi_NN2 and_CC its_APPGE block_NN1 are_VBR mapped_VVN to_II a_AT1 and_CC otherwise_RR the_AT block_NN1 is_VBZ mapped_VVN to_II b_ZZ1 ._. 
</s>
<s>
The_AT clone_NN1 host_NN1 location_NN1 may_VM affect_VVI the_AT network_NN1 latency_NN1 and_CC round_JJ trip_NN1 time_NNT1 ,_, which_DDQ may_VM have_VHI a_AT1 negative_JJ effect_NN1 on_II energy_NN1 efficiency_NN1 ._. 
</s>
<s>
Note_VV0 that_CST for_IF the_AT 256_MC MSN_NP1 experiments_NN2 ,_, the_AT best_JJT solution_NN1 found_VVN by_II different_JJ choice_NN1 of_IO parameters_NN2 has_VHZ ._. 
</s>
<s>
Advising_VVG individuals_NN2 to_TO exercise_VVI more_DAR is_VBZ a_AT1 straightforward_JJ approach_NN1 to_TO prevent_VVI and_CC treat_VVI obese_JJ and_CC overweight_JJ individuals_NN2 ._. 
</s>
<s>
If_CS A_ZZ1 satisfies_VVZ the_AT condition_NN1 in_II line_NN1 4_MC ,_, then_RT each_DD1 pair_NN 〈_NULL C_ZZ1 ,_, Di_NP1 〉_NULL is_VBZ removed_VVN from_II P_ZZ1 in_II line_NN1 5_MC thus_RR making_VVG P_ZZ1 |_NULL Cempty_JJ ._. 
</s>
<s>
Citation-based_JJ document_NN1 searches_NN2 are_VBR an_AT1 extremely_RR powerful_JJ tool_NN1 that_CST have_VH0 been_VBN shown_VVN to_TO provide_VVI results_NN2 which_DDQ complement_VV0 other_JJ document_NN1 retrieval_NN1 methods_NN2 ,_, such_II21 as_II22 keyword_NN1 searches_NN2 (_( Kuper_NP1 ,_, Nicholson_NP1 ,_, &;_NULL Hemingway_NP1 ,_, 2006_MC )_) ._. 
</s>
<s>
The_AT results_NN2 are_VBR reported_VVN in_II Fig._NN1 10_MC ._. 
</s>
<s>
Cmay_VV0 construct_VV0 a_AT1 delegation_NN1 attestation_NN1 of_IO its_APPGE declaration_NN1 of_IO UNAUTHENTICATED_JJ fory/kand_NN1 pass_VV0 that_DD1 attestation_NN1 to_II several_DA2 organizations_NN2 ._. 
</s>
<s>
Consequently_RR ,_, in_II the_AT continuing_JJ version_NN1 ,_, reward_NN1 received_VVN in_II one_MC1 "_" episode_NN1 "_" can_VM be_VBI attributed_VVN to_II state-action_JJ pairs_NN2 in_II the_AT previous_JJ "_" episode_NN1 "_" (_( and_CC farther_RRR in_II the_AT past_NN1 )_) ._. 
</s>
<s>
The_AT CGI_NP1 scripts_NN2 '_NULL main_JJ function_NN1 is_VBZ to_TO process_VVI user_NN1 input_NN1 and_CC performs_VVZ some_DD services_NN2 (_( such_II21 as_II22 retrieving_VVG data_NN from_II a_AT1 database_NN1 ,_, or_CC dynamically_RR creating_VVG a_AT1 web_NN1 page_NN1 )_) for_IF the_AT user_NN1 ._. 
</s>
<s>
To_II this_DD1 end_NN1 ,_, in_II future_JJ developments_NN2 of_IO this_DD1 software_NN1 we_PPIS2 aim_VV0 to_TO simulate_VVI allele_NN1 frequencies_NN2 and_CC how_RRQ they_PPHS2 change_VV0 over_II time_NNT1 ,_, including_II the_AT introduction_NN1 of_IO rare_JJ alleles_NN2 through_II dispersal_NN1 processes_NN2 and_CC thus_RR the_AT mixed_JJ effects_NN2 of_IO inbreeding_NN1 ,_, outbreeding_VVG and_CC mutation_NN1 ._. 
</s>
<s>
These_DD2 characteristics_NN2 distinguish_VV0 the_AT participatory_JJ modeling_NN1 approach_NN1 taken_VVN in_II the_AT St._NP1 Albans_NP1 Bay_NN1 watershed_NN1 from_II other_JJ methods_NN2 of_IO engaging_VVG the@public_FO in_II watershed_NN1 research_NN1 and_CC management_NN1 (_( Duram_NP1 and_CC Brown_NP1 ,_, 1999_MC )_) ._. 
</s>
<s>
Participants_NN2 were_VBDR not_XX formally_RR introduced_VVN to_II the_AT Inference_NN1 Inspector_NN1 prior_II21 to_II22 the_AT experiment_NN1 ,_, but_CCB some_DD of_IO the_AT basic_JJ functionality_NN1 was_VBDZ covered_VVN as_II part_NN1 of_IO the_AT preceding_JJ OWL_NN1 tutorial_NN1 ._. 
</s>
<s>
Other_JJ policies_NN2 of_IO these_DD2 types_NN2 (_( e.g._REX user_NN1 and_CC resource_VV0 restrictions_NN2 )_) can_VM be_VBI added_VVN as_CSA required_VVN by_II sites_NN2 ._. 
</s>
<s>
A_AT1 saved_JJ state_NN1 contains_VVZ all_DB the_AT information_NN1 needed_VVD to_TO resume_VVI a_AT1 simulation_NN1 starting_VVG from_II the_AT saved_JJ point_NN1 ,_, including_II all_DB the_AT model_NN1 parameter_NN1 values_NN2 and_CC variable_JJ values_NN2 for_IF the_AT duration_NN1 of_IO simulated_JJ evolution_NN1 ._. 
</s>
<s>
The_AT reason_NN1 for_IF this_DD1 might_VM also_RR be_VBI the_AT disregard_NN1 of_IO pervious_JJ catchment_NN1 area_NN1 in_II KAREN_NP1 ._. 
</s>
<s>
This_DD1 makes_VVZ interpretation_NN1 of_IO the_AT results_NN2 significantly_RR more_RGR transparent_JJ than_CSN would_VM be_VBI the_AT case_NN1 for_IF alternative_JJ '_NULL minimal_JJ '_NULL (_( or_CC even_RR fuzzy_JJ )_) control_VV0 strategies_NN2 ,_, because_CS the_AT control_NN1 architecture_NN1 will_VM be_VBI largely_RR determined_VVN by_II the_AT predicted_JJ dominant_JJ mode_NN1 dynamics_NN (_( see_VV0 Young_NP1 ,_, 1999_MC )_) of_IO the_AT climate_NN1 system_NN1 ._. 
</s>
<s>
Flux_NN1 redistribution_NN1 allows_VVZ only_JJ part_NN1 of_IO the_AT flux_NN1 to_TO be_VBI used_VVN to_TO compute_VVI the_AT solution_NN1 on_II a_AT1 small_JJ cell_NN1 ._. 
</s>
<s>
This_DD1 can_VM be_VBI represented_VVN logically_RR ,_, and_CC in_II the_AT network_NN1 model_NN1 N_ZZ1 ,_, by_II the_AT failure_NN1 of_IO t_ZZ1 links_VVZ on_II t_ZZ1 paths_NN2 ._. 
</s>
<s>
The_AT initial_JJ population_NN1 of_IO particles_NN2 is_VBZ chosen_VVN as_II a_AT1 subset_NN1 of_IO the_AT LHD_NP1 with_IW the_AT best_JJT objective_JJ function_NN1 values_NN2 ._. 
</s>
<s>
This_DD1 allows_VVZ to_TO identify_VVI shortcomings_NN2 in_II the_AT conceptual_JJ as_II31 well_II32 as_II33 procedural_JJ model_NN1 and_CC its_APPGE structure_NN1 ._. 
</s>
<s>
The_AT researchers_NN2 had_VHD to_TO learn_VVI how_RRQ to_TO implement_VVI models_NN2 using_VVG the_AT CCA_NP1 0.6.6_MC ,_, ESMF_NP1 3.1.1_MC and_CC OpenMI_NP1 1.4_MC frameworks_NN2 for_IF the_AT Thornthwaite_JJ model_NN1 implementations_NN2 ._. 
</s>
<s>
The_AT EBGM_NN1 method_NN1 uses_VVZ a_AT1 set_NN1 of_IO Gabor_NN1 jets_NN2 associated_VVN to_II nodes_NN2 that_CST extract_VV0 a_AT1 local_JJ feature_NN1 from_II the_AT face_NN1 ._. 
</s>
<s>
All_DB of_IO the_AT EMC_NN1 nodes_NN2 were_VBDR most_RGT sensitive_JJ to_II themselves_PPX2 ,_, flow_NN1 and_CC total_JJ suspended_JJ sediment_NN1 loads_NN2 in_II that_DD1 order_NN1 ._. 
</s>
<s>
In_II each_DD1 case_NN1 ,_, we_PPIS2 let_VV0 the_AT OMNI_NN1 converge_VVI before_II the_AT next_MD set_NN1 of_IO changes_NN2 is_VBZ effected_VVN ._. 
</s>
<s>
A_AT1 search_NN1 engine_NN1 that_CST is_VBZ good_JJ at_II finding_VVG questions_NN2 and_CC answers_VVZ in_II this_DD1 body_NN1 of_IO content_NN1 ._. 
</s>
<s>
According_II21 to_II22 the_AT results_NN2 of_IO the_AT experiment_NN1 ,_, the_AT impact_NN1 of_IO constraints_NN2 over_II NFPs_NP1 is_VBZ higher_JJR than_CSN the_AT impact_NN1 of_IO integrity_NN1 constraints_NN2 ._. 
</s>
<s>
We_PPIS2 show_VV0 experimental_JJ results_NN2 in_II Section_NN1 5_MC ._. 
</s>
<s>
The_AT correlation_NN1 degree_NN1 (_( based_VVN on_II the_AT proposed_JJ statistical_JJ correlation_NN1 measure_NN1 )_) between_II two_MC alerts_NN2 or_CC a_AT1 group_NN1 of_IO alerts_NN2 within_II a_AT1 candidate_NN1 frequent_JJ pattern_NN1 might_VM not_XX be_VBI significant_JJ at_II the_AT moment_NN1 ._. 
</s>
<s>
We_PPIS2 classified_VVD the_AT vulnerabilities_NN2 found_VVN by_II these_DD2 techniques_NN2 as_CSA either_RR implementation_NN1 bugs_NN2 or_CC design_NN1 flaws_NN2 ._. 
</s>
<s>
At_RR21 worst_RR22 one_PN1 may_VM not_XX even_RR be_VBI able_JK to_TO execute_VVI f_ZZ1 in_II multiple_JJ concurrent_JJ threads_NN2 ._. 
</s>
<s>
A_AT1 more_RGR principled_JJ way_NN1 of_IO combining_VVG our_APPGE different_JJ features_NN2 for_IF computing_VVG the_AT intent-ness_JJ score_NN1 remains_VVZ an_AT1 important_JJ future_JJ work_NN1 ._. 
</s>
<s>
Nevertheless_RR ,_, as_CSA we_PPIS2 show_VV0 below_II the_AT performance_NN1 is_VBZ further_RRR boosted_VVN when_CS the_AT two_MC methods_NN2 are_VBR combined_VVN ._. 
</s>
<s>
Veiled_JJ certificates_NN2 utilize_VV0 existing_JJ technology_NN1 and_CC can_VM be_VBI implemented_VVN within_II the_AT current_JJ standards_NN2 and_CC public_JJ key_JJ infrastructure_NN1 ._. 
</s>
<s>
In_II adaptive_JJ integration_NN1 schemes_NN2 &lsqb;_( 48_MC &rsqb;_) ,_, the_AT point-singularity_NN1 that_CST lies_VVZ inside_II the_AT finite_JJ element_NN1 is_VBZ resolved_VVN by_II recursively_RR subdividing_VVG the_AT element_NN1 into_II nonconforming_VVG subcells_NN2 until_CS a_AT1 prescribed_JJ error_NN1 tolerance_NN1 is_VBZ achieved_VVN ._. 
</s>
<s>
In_II Fig._NN1 7(a)_FO we_PPIS2 plot_VV0 the_AT Hermite_JJ cubic_JJ collocation_NN1 fracture_NN1 half-widths_NN2 h(x)_NNU for_IF the_AT following_JJ sequence_NN1 of_IO values_NN2 for_IF the_AT fracture_NN1 toughness_NN1 K_ZZ1 =_FO 1_MC1 ,_, 2_MC ,_, 4_MC ,_, 8_MC ,_, 16_MC ,_, and_CC 32_MC ._. 
</s>
<s>
This_DD1 starts_VVZ after_II the_AT start_NN1 of_IO the_AT server_NN1 transfer_NN1 time_NNT1 and_CC finishes_VVZ after_II the_AT end_NN1 of_IO the_AT server_NN1 transfer_NN1 time_NNT1 ._. 
</s>
<s>
Three_MC currently_RR defined_VVN roles_NN2 exist_VV0 in_II the_AT system_NN1 ._. 
</s>
<s>
A_AT1 step-by-step_JJ SPM_NP1 composition_NN1 algorithm_NN1 is_VBZ shown_VVN in_II Fig._NN1 13_MC ._. 
</s>
<s>
The_AT framework_NN1 developed_VVN in_II this_DD1 study_NN1 was_VBDZ aimed_VVN at_II providing_VVG environmental_JJ managers_NN2 a_AT1 set_NN1 of_IO simplified_JJ tools_NN2 to_TO evaluate_VVI changes_NN2 in_II coastal_JJ habitats_NN2 due_II21 to_II22 these_DD2 important_JJ processes_NN2 ._. 
</s>
<s>
Similarly_RR ,_, McPAD_NP1 (_( Perdisci_NP1 et_RA21 al._RA22 ,_, 2009_MC )_) creates_VVZ 2v-grams_NNU2 and_CC uses_VVZ a_AT1 sliding_JJ window_NN1 to_TO cover_VVI all_DB sets_NN2 of_IO 2_MC bytes_NNU2 ,_, positions_NN2 apart_RL in_II network_NN1 traffic_NN1 payloads_NN2 ._. 
</s>
<s>
Firstly_RR ,_, an_AT1 SD_NP1 approach_NN1 is_VBZ widely_RR recognised_VVN as_II a_AT1 clear_JJ method_NN1 for_IF communicating_JJ ideas_NN2 and_CC complex_JJ structures_NN2 to_II those_DD2 with_IW little_DA1 working_JJ knowledge_NN1 of_IO the_AT particular_JJ problem_NN1 studied_VVD ._. 
</s>
<s>
Learning_VVG to_TO ask_VVI the_AT right_JJ questions_NN2 is_VBZ also_RR important_JJ "_" since_CS you_PPY have_VH0 to_TO have_VHI a_AT1 great_JJ crap_NN1 detector_NN1 when_CS it_PPH1 comes_VVZ to_II technology_NN1 "_" ._. 
</s>
<s>
This_DD1 is_VBZ a_AT1 common_JJ choice_NN1 in_II the_AT study_NN1 of_IO mean-field_JJ models_NN2 ,_, and_CC is_VBZ partially_RR justified_VVN by_II the_AT observation_NN1 ,_, that_CST each_DD1 part_NN1 of_IO the_AT cortex_NN1 is_VBZ connected_VVN to_II each_PPX221 other_PPX222 part_NN1 ._. 
</s>
<s>
The_AT combinations_NN2 are_VBR defined_VVN by_II selecting_VVG the_AT POD_NN1 coefficients_NN2 using_VVG a_AT1 DOE_NN1 ._. 
</s>
<s>
Mobile_JJ communications_NN2 between_II mobile_JJ devices_NN2 and_CC base_NN1 stations_NN2 are_VBR governed_VVN by_II various_JJ protocols_NN2 and_CC standards_NN2 such_II21 as_II22 those_DD2 defined_VVN by_II 3GPP_FO and_CC 3GPP2_FO ._. 
</s>
<s>
As_II an_AT1 architectural_JJ consideration_NN1 ,_, if_CS authorization_NN1 is_VBZ to_TO occur_VVI over_RP multiple_JJ domains_NN2 ,_, a_AT1 global_JJ namespace_NN1 for_IF subjects_NN2 ,_, objects_NN2 ,_, etc._RA ,_, may_VM be_VBI required_VVN ._. 
</s>
<s>
These_DD2 systems_NN2 can_VM be_VBI enhanced_VVN further_RRR by_II NLG_NP1 components_NN2 capable_JJ of_IO generating_JJ responses_NN2 that_CST better_JJR address_NN1 the_AT users_NN2 '_NULL questions_NN2 &lsqb;_( 1_MC1 &rsqb;_) ._. 
</s>
<s>
Overview_NN1 of_IO the_AT LE-DAnCE_JJ locality_NN1 manager_NN1 ._. 
</s>
<s>
Table_NN1 1_MC1 provides_VVZ a_AT1 list_NN1 of_IO independent_JJ and_CC dependent_JJ variables_NN2 ._. 
</s>
<s>
However_RR ,_, it_PPH1 still_RR has_VHZ many_DA2 unavoidable_JJ drawbacks_NN2 ,_, for_REX21 example_REX22 ,_, there_EX is_VBZ a_AT1 rather_RG large_JJ number_NN1 of_IO parameters_NN2 need_VV0 to_TO be_VBI manually_RR fixed_VVN or_CC estimated_VVD ;_; heavy_JJ burden_NN1 of_IO memory_NN1 consuming_VVG and_CC computation_NN1 time_NNT1 for_IF the_AT calculation_NN1 of_IO convolution_NN1 ;_; along_II21 with_II22 lots_PN of_IO redundancy_NN1 when_CS extracting_VVG orientation_NN1 information_NN1 &lsqb;_( 13_MC &rsqb;_) ._. 
</s>
<s>
Another_DD1 experiment_NN1 was_VBDZ performed_VVN with_IW galaxies_NN2 taken_VVN from_II the_AT Cassowary_JJ catalog_NN1 &lsqb;_( 34_MC &rsqb;_) ,_, which_DDQ are_VBR gravitational_JJ lens_NN1 systems_NN2 detected_VVN in_II the_AT data_NN releases_NN2 of_IO SDSS_NP1 ._. 
</s>
<s>
The_AT ICARUS_NP1 framework_NN1 is_VBZ expressive_JJ enough_RR to_TO enable_VVI all_DB existing_JJ MCTS_NP1 enhancements_NN2 to_TO be_VBI defined_VVN and_CC provides_VVZ a_AT1 new_JJ tool_NN1 for_IF discovery_NN1 of_IO novel_JJ types_NN2 of_IO enhancement_NN1 ._. 
</s>
<s>
As_CSA will_VM become_VVI clearer_JJR in_II Section_NN1 6.5_MC ,_, this_DD1 choice_NN1 was_VBDZ made_VVN after_II some_DD preliminary_JJ experiments_NN2 that_CST involved_VVD more_RGR complex_JJ cell_NN1 functions_NN2 ._. 
</s>
<s>
Cluster_VV0 C3In_FO order_NN1 to_TO allow_VVI the_AT users_NN2 to_TO make_VVI use_NN1 of_IO the_AT benefits_NN2 of_IO "_" data-browsing_NN1 interfaces_VVZ "_" a_AT1 sufficient_JJ supply_NN1 of_IO heterogeneous_JJ data_NN is_VBZ required_VVN ;_; otherwise_RR the_AT system_NN1 can_VM not_XX best_RRT a_AT1 traditional_JJ solution_NN1 ._. 
</s>
<s>
Finally_RR ,_, we_PPIS2 also_RR compared_VVD the_AT production_NN1 ,_, in_II the_AT absence_NN1 of_IO any_DD management_NN1 ,_, of_IO the_AT two_MC Mediterranean_JJ sites_NN2 with_IW two_MC imaginary_JJ sites_NN2 with_IW the_AT same_DA physical_JJ features_NN2 but_CCB located_VVD in_II northern_JJ Europe_NP1 (_( i.e._REX changing_VVG the_AT location_NN1 from_II "_" Mediterranean_JJ basin_NN1 "_" to_II "_" North_ND1 and_CC Baltic_JJ seas_NN2 "_" and_CC setting_VVG the_AT same_DA glass_NN1 eel_NN recruitment_NN1 through_II the_AT advanced_JJ interface_NN1 )_) ,_, thus_RR considering_VVG different_JJ body_NN1 growth_NN1 ,_, maximum_JJ settlement_NN1 potential_NN1 and_CC mortality_NN1 ,_, in_BCL21 order_BCL22 to_TO have_VHI some_DD suggestions_NN2 regarding_II the_AT so-called_JJ restocking_NN1 dilemma_NN1 (_( i.e._REX whether_CSW it_PPH1 is_VBZ better_JJR to_TO restock_VVI in_II a_AT1 Mediterranean_JJ site_NN1 ,_, where_CS eels_NN2 grow_VV0 faster_JJR but_CCB smaller_JJR ,_, mortality_NN1 is_VBZ higher_JJR and_CC productivity_NN1 too_RR ,_, or_CC in_II a_AT1 site_NN1 of_IO Northern_JJ Europe_NP1 ,_, where_CS eels_NN2 grow_VV0 more_RGR slowly_RR but_CCB larger_JJR ,_, mortality_NN1 is_VBZ lower_JJR and_CC so_RR is_VBZ productivity_NN1 )_) ._. 
</s>
<s>
Active_JJ cyber_NN1 defenses_NN2 should_VM respect_VVI the_AT civil_JJ liberties_NN2 of_IO all_DB persons_NN2 affected_VVN ,_, including_II their_APPGE rights_NN2 of_IO privacy_NN1 ,_, free_JJ speech_NN1 ,_, and_CC association_NN1 ._. 
</s>
<s>
Additionally_RR Expert7_FO reiterates_VVZ the_AT importance_NN1 about_II including_VVG Forensic_JJ Stakeholders_NN2 in_II the_AT development_NN1 process_NN1 "_" As_CSA staff_NN who_PNQS understand_VV0 the_AT scope_NN1 and_CC difficulty_NN1 of_IO the_AT forensic_JJ efforts_NN2 ,_, the_AT tech_NN1 &lsqb;_( forensic_JJ &rsqb;_) group_NN1 should_VM work_VVI closely_RR with_IW the_AT policy_NN1 makers_NN2 "_" ._. 
</s>
<s>
Indeed_RR an_AT1 agent_NN1 could_VM reason_VVI about_II the_AT relations_NN2 between_II their_APPGE own_DA messages_NN2 when_CS matching_VVG a_AT1 received_JJ one_PN1 ._. 
</s>
<s>
This_DD1 is_VBZ any_DD consideration_NN1 given_VVN to_II the_AT ability_NN1 of_IO the_AT software_NN1 to_TO perform_VVI new_JJ tasks_NN2 ,_, and_CC cope_VVI with_IW any_DD changes_NN2 to_II existing_JJ tasks_NN2 at_II run_NN1 time_NNT1 ._. 
</s>
<s>
Let_VV0 then_RT @DDGk_FO denote_VVI the_AT operator_NN1 that_CST uses_VVZ @k_FO for_IF the_AT computation_NN1 of_IO the_AT numerical_JJ traces_NN2 ._. 
</s>
<s>
In_II this_DD1 table_NN1 userid_NN1 for_IF both_DB2 users_NN2 being_VBG friends_NN2 are_VBR inserted_VVN in_II fields_NN2 '_NULL user1id_FO and_CC user2id_FO ._. 
</s>
<s>
This_DD1 duplication_NN1 results_NN2 in_II a_AT1 highly_RR periodic_JJ dataset_NN1 ._. 
</s>
<s>
Global_JJ set-up_NN1 parameters_NN2 are_VBR for_REX21 example_REX22 the_AT switches_NN2 for_IF those_DD2 steps_NN2 that_CST can_VM be_VBI turned_VVN off_RP ,_, and_CC periodicities_NN2 the_AT user_NN1 chooses_VVZ for_IF some_DD steps_NN2 ._. 
</s>
<s>
Circles_NN2 indicate_VV0 the_AT extent_NN1 to_II which_DDQ each_DD1 process_NN1 quality_NN1 attribute_NN1 was_VBDZ identified_VVN by_II each_DD1 paper_NN1 ._. 
</s>
<s>
As_CSA mentioned_VVN earlier_RRR ,_, there_EX are_VBR 150·n_FO iterations_NN2 required_VVN to_TO estimate_VVI the_AT dissimilarity_NN1 thresholds_NN2 for_IF n_ZZ1 items_NN2 ._. 
</s>
<s>
Strategy_NN1 2_MC being_VBG retained_VVN as_II the_AT most_RGT efficient_JJ ,_, two_MC naive_JJ users_NN2 (_( non-familiar_JJ with_IW evolutionary_JJ algorithms_NN2 nor_CC with_IW image_NN1 processing_NN1 )_) have_VH0 tested_VVN IPAT_NN1 ._. 
</s>
<s>
Given_VVN this_DD1 variability_NN1 the_AT BBN_NP1 may_VM still_RR be_VBI useful_JJ in_II providing_VVG reasonable_JJ estimates_NN2 of_IO the_AT likely_JJ outcomes_NN2 of_IO planned_JJ interventions_NN2 ._. 
</s>
<s>
These_DD2 malicious_JJ nodes_NN2 have_VH0 the_AT ability_NN1 to_TO exploit_VVI the_AT threat_NN1 that_CST there_EX is_VBZ no_AT assurance_NN1 between_II the_AT binding_NN1 of_IO the_AT node_NN1 identity_NN1 and_CC the_AT public_JJ key_NN1 ._. 
</s>
<s>
Otherwise_RR ,_, the_AT pixel_NN1 in_II the_AT machine-detected_JJ object_NN1 is_VBZ counted_VVN as_II a_AT1 false_JJ positive_JJ (_( FP_NP1 )_) ._. 
</s>
<s>
These_DD2 are_VBR not_XX the_AT only_JJ event_NN1 tracking_NN1 approaches_VVZ that_CST can_VM be_VBI used_VVN in_II transmedia_NN1 learning_NN1 campaigns_NN2 and_CC readers_NN2 are_VBR encouraged_VVN to_TO reflect_VVI on_II how_RRQ their_APPGE own_DA computational_JJ approaches_NN2 might_VM also_RR apply_VVI ._. 
</s>
<s>
The_AT conjunction_NN1 years_NNT2 are_VBR most_RGT related_JJ ,_, which_DDQ is_VBZ the_AT same_DA to_II our_APPGE common_JJ sense_NN1 ._. 
</s>
<s>
In_II this_DD1 context_NN1 ,_, we_PPIS2 briefly_RR overview_NN1 previous_JJ findings_NN2 on_II the_AT shortcomings_NN2 of_IO the_AT available_JJ AOP_NN1 languages_NN2 for_IF security_NN1 (_( AlHadidi_NP1 etal._NNU ,_, 2006_MC )_) ,_, particularly_RR in_II AspectJ_NP1 ,_, and_CC we_PPIS2 present_VV0 some_DD possible_JJ extensions_NN2 to_II AspectJ_NP1 in_BCL21 order_BCL22 to_TO handle_VVI security_NN1 issues_NN2 ._. 
</s>
<s>
As_II31 well_II32 as_II33 our_APPGE proposal_NN1 RCU_NN1 allows_VVZ simultaneous_JJ access_NN1 between_II a_AT1 single_JJ write_VV0 transaction_NN1 and_CC multiple_NN1 read_VVD transactions_NN2 ._. 
</s>
<s>
In_II Section_NN1 2_MC we_PPIS2 give_VV0 some_DD background_NN1 concerning_II coalitional_JJ games_NN2 ,_, and_CC define_VV0 the_AT CSG_NP1 model_NN1 ._. 
</s>
<s>
First_MD ,_, the_AT provenance_NN1 would_VM still_RR need_VVI to_TO refer_VVI to_II the_AT identified_JJ resources_NN2 of_IO which_DDQ people_NN wish_VV0 to_TO describe_VVI the_AT provenance_NN1 ,_, e.g._REX a_AT1 Web_NN1 page_NN1 identified_VVN by_II its_APPGE URI_NN2 ,_, and_CC these_DD2 are_VBR mutable_JJ ._. 
</s>
<s>
We_PPIS2 believe_VV0 that_CST this_DD1 justifies_VVZ the_AT adoption_NN1 of_IO a_AT1 mixed_JJ approach_NN1 to_TO rank_VVI entity_NN1 types_NN2 and_CC partly_RR explains_VVZ the_AT high_JJ effectiveness_NN1 achieved_VVN by_II the_AT DEC-TREE_JJ approach_NN1 ._. 
</s>
<s>
This_DD1 step_NN1 must_VM be_VBI carried_VVN out_RP before_II any_DD other_JJ SIP_NN1 messages_NN2 are_VBR exchanged_VVN ._. 
</s>
<s>
Adapted_VVN from_II Carstensen_NP1 et_RA21 al._RA22 ,_, 1997_MC ,_, Dochain_JJ and_CC Vanrolleghem_NP1 ,_, 2001_MC ,_, Jakeman_NP1 et_RA21 al._RA22 ,_, 2006_MC ,_, and_CC Refsgaard_NP1 et_RA21 al_RA22 ._. 
</s>
<s>
(_( 2007_MC )_) ._. 
</s>
<s>
It_PPH1 also_RR includes_VVZ the_AT command_NN1 setPolicies_NN2 ._. 
</s>
<s>
The_AT DCO_NN1 B_ZZ1 &B;_NULL algorithm_NN1 ,_, by_II design_NN1 ,_, is_VBZ guaranteed_VVN to_TO reproduce_VVI the_AT model_NN1 selection_NN1 results_VVZ that_CST would_VM be_VBI obtained_VVN using_VVG an_AT1 exhaustive_JJ search_NN1 of_IO possible_JJ models_NN2 based_VVN on_II a_AT1 set_NN1 of_IO candidate_NN1 covariates_NN2 using_VVG a_AT1 given_JJ model_NN1 selection_NN1 approach_NN1 (_( BIC_NP1 in_II our_APPGE case_NN1 )_) ._. 
</s>
<s>
Indeed_RR ,_, adding_VVG a_AT1 new_JJ reviewer_NN1 to_II a_AT1 Document_NN1 should_VM not_XX influence_VVI the_AT contents_NN2 of_IO a_AT1 review_NN1 being_VBG entered_VVN concurrently_RR ._. 
</s>
<s>
For_IF the_AT DirectSubtrees_NN2 kernel_VV0 we_PPIS2 use_VV0 the_AT second_MD definition_NN1 of_IO Freq_NP1 ._. 
</s>
<s>
We_PPIS2 selected_VVD Wednesday_NPD1 ,_, 28/05/2014_MF for_IF the_AT manual_JJ evaluation_NN1 of_IO clusters_NN2 and_CC cross-links_NN2 ._. 
</s>
<s>
The_AT hunter_NN1 has_VHZ five_MC possible_JJ actions_NN2 ;_; move_VV0 ?_? north_ND1 ,_, ?_? south_ND1 ,_, ?_? east_ND1 ,_, or_CC ?_? west_ND1 ,_, and_CC ?_? catch_NN1 ?_? (_( the_AT latter_DA is_VBZ applicable_JJ only_JJ when_CS the_AT hunter_NN1 and_CC prey_NN1 are_VBR in_II the_AT same_DA location_NN1 )_) ._. 
</s>
<s>
A_AT1 hybrid_JJ algorithm_NN1 for_IF the_AT recognition_NN1 of_IO classes_NN2 of_IO objects_NN2 and_CC concepts_NN2 in_II images_NN2 has_VHZ been_VBN developed_VVN in_II &lsqb;_( 13_MC &rsqb;_) where_RRQ a_AT1 generative_JJ model_NN1 is_VBZ used_VVN to_TO create_VVI a_AT1 fixed_JJ length_NN1 representation_NN1 of_IO the_AT image_NN1 ,_, which_DDQ can_VM be_VBI classified_VVN then_RT using_VVG a_AT1 discriminative_JJ technique_NN1 ._. 
</s>
<s>
For_IF every_AT1 matching_JJ document_NN1 we_PPIS2 compute_VV0 counts_NN2 of_IO hits_NN2 of_IO different_JJ types_NN2 at_II different_JJ proximity_NN1 levels_NN2 ._. 
</s>
<s>
Parameters_NN2 used_VVN for_IF those_DD2 two_MC benefit_NN1 functions_NN2 came_VVD from_II empirical_JJ analysis_NN1 of_IO urban_JJ water_NN1 use_NN1 patterns_NN2 under_II varying_JJ water_NN1 prices_NN2 along_II21 with_II22 data_NN from_II farm_NN1 enterprise_NN1 budgets_NN2 adjusted_VVN to_II various_JJ water_NN1 supply_NN1 conditions_NN2 ._. 
</s>
<s>
Adopting_VVG a_AT1 forensic_JJ readiness_NN1 plan_NN1 is_VBZ also_RR likely_JJ to_TO improve_VVI the_AT security_NN1 situation_NN1 awareness_NN1 of_IO the_AT organization_NN1 (_( see_VV0 Webb_NP1 et_RA21 al._RA22 ,_, 2014_MC )_) ._. 
</s>
<s>
We_PPIS2 consider_VV0 vehicle_NN1 tracking_NN1 based_VVN on_II the_AT distinctive_JJ acoustic_JJ signatures_NN2 produced_VVN by_II different_JJ vehicles_NN2 ._. 
</s>
<s>
After_CS refinement_NN1 ,_, a_AT1 new_JJ end-triangle_NN1 is_VBZ produced_VVN ,_, that_DD1 represents_VVZ the_AT end_NN1 of_IO the_AT stroke_NN1 much_RR better_RRR ._. 
</s>
<s>
ROs_NP1 enable_VV0 initial_JJ investments_NN2 to_TO be_VBI reduced_VVN by_II postponing_VVG some_DD decisions_NN2 for_IF the_AT future_NN1 ._. 
</s>
<s>
When_CS calculating_VVG such_DA risk_NN1 values_NN2 ,_, we_PPIS2 include_VV0 the_AT risk_NN1 of_IO the_AT permissions_NN2 that_CST can_VM be_VBI explicitly_RR acquired_VVN through_II those_DD2 roles_NN2 ,_, as_II31 well_II32 as_II33 those_DD2 that_CST can_VM be_VBI inferred_VVN from_II them_PPHO2 ._. 
</s>
<s>
Note_VV0 that_CST it_PPH1 is_VBZ not_XX possible_JJ to_TO obtain_VVI real_JJ measures_NN2 of_IO security_NN1 effectiveness_NN1 across_II organizations_NN2 ._. 
</s>
<s>
Moreover_RR ,_, betweenness_NN1 plots_NN2 in_II Section4indicated_FO that_CST there_EX are_VBR a_AT1 few_DA2 overlay_VV0 nodes_NN2 that_CST are_VBR key_JJ to_II most_DAT of_IO the_AT overlay_NN1 paths_NN2 ._. 
</s>
<s>
Multilevel_VV0 resilience_NN1 is_VBZ defined_VVN with_II31 respect_II32 to_II33 protocol_NN1 layer_NN1 ,_, protocol_NN1 plane_NN1 ,_, and_CC hierarchical_JJ network_NN1 organisation_NN1 &lsqb;_( 97_MC &rsqb;_) ,_, &lsqb;_( 143_MC &rsqb;_) and&lsqb;144&rsqb;_FO ._. 
</s>
<s>
The_AT most_RGT common_JJ approach_NN1 to_II calculating_VVG optimal_JJ flows_NN2 in_II such_DA networks_NN2 is_VBZ the_AT one_PN1 we_PPIS2 adopt_VV0 here_RL ,_, i.e._REX ,_, creating_VVG and_CC doing_VDG calculations_NN2 on_II a_AT1 graph_NN1 comprised_VVN of_IO a_AT1 cascade_NN1 of_IO subgraphs_NN2 ,_, each_DD1 one_PN1 related_VVD to_II a_AT1 different_JJ epoch_NN1 ,_, called_VVN time-expanded_JJ graph_NN1 ._. 
</s>
<s>
In_II our_APPGE roles_NN2 as_CSA analysts_NN2 ,_, we_PPIS2 studied_VVD the_AT diagrams_NN2 shown_VVN in_II Fig._NN1 9_MC and_CC Fig._NN1 11_MC ,_, and_CC it_PPH1 became_VVD apparent_JJ that_CST there_EX were_VBDR potentially_RR missing_JJ Rules_NN2 regarding_II the_AT HOT_JJ self-survey_NN1 ._. 
</s>
<s>
Five_MC such_DA tables_NN2 would_VM give_VVI an_AT1 overall_JJ probability_NN1 of_IO 99.9%_FO ._. 
</s>
<s>
Since_CS the_AT performance_NN1 of_IO the_AT algorithm_NN1 depends_VVZ on_II the_AT number_NN1 of_IO reported_JJ frequent_JJ structures_NN2 ,_, we_PPIS2 exclude_VV0 the_AT trials_NN2 that_CST returned_VVD significantly_RR more_RGR frequent_JJ patterns_NN2 compared_VVN to_II the_AT average_JJ number_NN1 of_IO frequent_JJ patterns_NN2 ._. 
</s>
<s>
For_IF each_DD1 year_NNT1 j_ZZ1 within_II the_AT period_NN1 ,_, the_AT ratio_NN1 between_II the_AT average_JJ ring_NN1 width_NN1 (_( radial_JJ increment_NN1 )_) of_IO year_NNT1 j_ZZ1 and_CC the_AT average_JJ multi-year_JJ radial_JJ increment_NN1 for_IF the_AT whole_JJ period_NN1 was_VBDZ estimated_VVN using_VVG the_AT increment_NN1 core_NN1 data_NN ._. 
</s>
<s>
The_AT candidates_NN2 with_IW the_AT lowest_JJT Borda_NN1 score_NN1 before_II the_AT manipulator_NN1 '_NULL s_ZZ1 vote_NN1 are_VBR p_ZZ1 followed_VVN by_II all_DB aj_VV0 '_NULL s_ZZ1 ,_, which_DDQ all_DB have_VH0 1_MC1 more_DAR ,_, as_CSA explained_VVN above_RL ._. 
</s>
<s>
There_EX is_VBZ a_AT1 difference_NN1 between_II the_AT classification_NN1 of_IO outliers_NN2 and_CC events_NN2 ,_, as_CSA a_AT1 succession_NN1 of_IO outliers_NN2 is_VBZ surely_RR stronger_JJR evidence_NN1 to_II an_AT1 event_NN1 occurrence_NN1 than_CSN a_AT1 single_JJ one_PN1 ._. 
</s>
<s>
We_PPIS2 also_RR conclude_VV0 that_CST that_DD1 the_AT information_NN1 retrieval_NN1 subfield_NN1 '_NULL s_ZZ1 body_NN1 of_IO literature_NN1 is_VBZ expanding_VVG into_II areas_NN2 not_XX extensively_RR covered_VVN during_II the_AT years_NNT2 prior_II21 to_II22 2000_MC ._. 
</s>
<s>
Otherwise_RR put_VVN in_II some_DD identification_NN1 of_IO your_APPGE project_NN1 ._. 
</s>
<s>
Fig._NN1 @3_FO )_) :_: carry_VV0 out_RP ANOVA_NN1 to_TO test_VVI for_IF differences_NN2 between_II areas_NN2 ,_, and_CC calculate_VV0 means_NN and_CC standard_JJ deviations_NN2 for_IF each_DD1 key_JJ variable_NN1 for_IF both_DB2 areas_NN2 (_( reference_NN1 and_CC exposure_NN1 )_) ._. 
</s>
<s>
For_REX21 example_REX22 ,_, lawyers_NN2 who_PNQS appraise_VV0 policy_NN1 claims_NN2 have_VH0 to_TO use_VVI the_AT evidence_NN1 which_DDQ they_PPHS2 are_VBR given_VVN to_TO support_VVI or_CC refute_VVI the_AT claim_NN1 ._. 
</s>
<s>
Both_DB2 join_VV0 operations_NN2 will_VM compute_VVI the_AT resulting_JJ localized_JJ variable_NN1 binding_VVG 3l_NNU that_DD1 is_VBZ also_RR known_VVN by_II both_DB2 compute_VV0 nodes_NN2 ._. 
</s>
<s>
In_II static_JJ multirate_JJ case_NN1 ,_, each_DD1 station_NN1 selects_VVZ a_AT1 specific_JJ rate_NN1 at_II association_NN1 time_NNT1 and_CC may_VM change_VVI the_AT rate_NN1 occasionally_RR ,_, but_CCB not_XX very_RG often_RR ._. 
</s>
<s>
A_AT1 similar_JJ argument_NN1 applies_VVZ to_II eventually_RR periodic_JJ sequences_NN2 ,_, so_CS no_AT eventually_RR periodic_JJ sequence_NN1 has_VHZ a_AT1 bounded_JJ discrepancy_NN1 ._. 
</s>
<s>
In_II the_AT test_NN1 dataset_NN1 ,_, Virut_NP1 is_VBZ detected_VVN 194_MC times_NNT2 with_IW 52_MC sessions_NNT2 to_II live.com_NNU ,_, and_CC Sogou_NP1 12_MC times_NNT2 with_IW all_DB traffic_NN1 to_II sogou.com_NNU ._. 
</s>
<s>
However_RR ,_, no_AT such_DA controls_NN2 exist_VV0 with_II31 respect_II32 to_II33 users_NN2 ,_, as_CSA this_DD1 would_VM present_VVI a_AT1 barrier_NN1 and_CC inhibit_VVI network_NN1 externalities_NN2 ._. 
</s>
<s>
After_CS the_AT knowledge_NN1 base_NN1 was_VBDZ built_VVN ,_, during_II the_AT evaluation_NN1 phase_NN1 ,_, There_EX is_VBZ surprisingly_RR little_RR empirical_JJ software_NN1 engineering_NN1 research_NN1 that_CST has_VHZ analysed_VVN and_CC reported_JJ project_NN1 behaviour_NN1 at_II the_AT level_NN1 of_IO the_AT project_NN1 and_CC over_II time_NNT1 ._. 
</s>
<s>
In_II this_DD1 way_NN1 ,_, entailment_NN1 regimes_NN2 only_RR modify_VV0 the_AT evaluation_NN1 of_IO BGPs_NP1 but_CCB not_XX that_DD1 of_IO other_JJ SPARQL_NN1 operators_NN2 ._. 
</s>
<s>
If_CS there_EX are_VBR any_DD warnings_NN2 to_TO be_VBI reported_VVN ,_, the_AT element_NN1 names_NN2 in_II the_AT reduced_JJ net_NN1 are_VBR mapped_VVN back_RP to_II the_AT names_NN2 of_IO tasks_NN2 and_CC conditions_NN2 in_II the_AT original_JJ net_NN1 ._. 
</s>
<s>
This_DD1 is_VBZ an_AT1 offline_JJ recovery_NN1 procedure_NN1 ,_, which_DDQ works_VVZ only_RR for_IF rootkits_NN2 that_CST modify_VV0 the_AT system_NN1 call_NN1 table_NN1 ._. 
</s>
<s>
Recall_VV0 that_CST also_RR occurs_VVZ exactly_RR once_RR ,_, but_II21 for_II22 an_AT1 URI_NN1 that_CST has_VHZ three_MC possible_JJ values_NN2 for_IF dct:subject_FO ,_, and_CC so_RR ,_, it_PPH1 receives_VVZ weight_NN1 13_MC 15=115_FO ._. 
</s>
<s>
The_AT number_NN1 of_IO bytes_NNU2 sent_VVN out_RP per_II core_NN1 per_II step_NN1 (_( S<x>_FO )_) is_VBZ dependent_JJ on_II the_AT geometry_NN1 used_VVN as_II31 well_II32 as_II33 the_AT number_NN1 of_IO sites_NN2 per_II core_NN1 ._. 
</s>
<s>
For_IF correct_JJ alignment_NN1 ,_, the_AT images_NN2 have_VH0 similar_JJ curvature_NN1 information_NN1 in_II overlapping_JJ regions_NN2 ,_, and_CC it_PPH1 will_VM output_VVI high_JJ global_JJ normalized_JJ correlation_NN1 coefficient_NN1 ._. 
</s>
<s>
Because_CS specific_JJ function_NN1 calls_NN2 are_VBR caught_VVN and_CC manipulated_VVN at_II runtime_NNT1 to_TO change_VVI their_APPGE original_JJ behavior_NN1 ,_, FCI_NP1 enables_VVZ transparent_JJ code_NN1 modification_NN1 without_IW directly_RR rewriting_VVG the_AT existing_JJ code_NN1 ._. 
</s>
<s>
Despite_II the_AT fact_NN1 that_CST these_DD2 headers_NN2 can_VM easily_RR be_VBI spoofed_VVN ,_, when_CS operating_VVG in_II combination_NN1 with_IW a_AT1 Bayesian_JJ filter_NN1 ,_, overall_JJ accuracy_NN1 is_VBZ approximately_RR doubled_VVN ._. 
</s>
<s>
Note_VV0 that_CST ,_, in_II baseline_NN1 FR_NP1 ,_, all_DB training_NN1 ,_, gallery_NN1 ,_, and_CC probe_VV0 images_NN2 are_VBR uncompressed_JJ ,_, so_CS21 that_CS22 no_AT mismatch_NN1 is_VBZ possible_JJ in_II31 terms_II32 of_II33 compression_NN1 ratio_NN1 ._. 
</s>
<s>
REST_VV0 '_NULL s_ZZ1 simplicity_NN1 ,_, along_II21 with_II22 its_APPGE natural_JJ fit_NN1 over_II HTTP_NP1 ,_, has_VHZ contributed_VVN to_II its_APPGE status_NN1 as_II a_AT1 method_NN1 of_IO choice_NN1 for_IF Web_NN1 2.0_MC applications_NN2 to_TO expose_VVI their_APPGE data_NN ._. 
</s>
<s>
Also_RR ,_, the_AT output_NN1 suffers_VVZ from_II noise_NN1 ._. 
</s>
<s>
If_CS an_AT1 organization_NN1 chooses_VVZ to_TO use_VVI a_AT1 prefix_NN1 of_IO addresses_NN2 under_II its_APPGE ownership_NN1 for_IF its_APPGE own_DA hosts_NN2 ,_, rather_II21 than_II22 delegating_VVG the_AT ownership_NN1 of_IO the_AT prefix_NN1 to_II another_DD1 organization_NN1 ,_, it_PPH1 will_VM assign_VVI that_DD1 prefix_NN1 of_IO addresses_NN2 to_II one_MC1 of_IO its_APPGE ASes_NP2 ._. 
</s>
<s>
Accordingly_RR we_PPIS2 define_VV0 lmaxi=l0i×x0_FO and_CC A0jj=1_FO ..._... 
</s>
<s>
Nt_NP1 for_IF each_DD1 triangular_JJ plaquette_NN1 ._. 
</s>
<s>
Instead_RR ,_, it_PPH1 is_VBZ caused_VVN by_II the_AT higher_JJR number_NN1 of_IO connections_NN2 within_II one_MC1 graph_NN1 chunk_NN1 leading_VVG to_II higher_JJR graph_NN1 chunk_NN1 diameters_NN2 ._. 
</s>
<s>
In_II this_DD1 work_NN1 a_AT1 simple_JJ water_NN1 network_NN1 it_PPH1 is_VBZ used_VVN to_TO illustrate_VVI the_AT approach_NN1 ._. 
</s>
<s>
The_AT third_MD contribution_NN1 is_VBZ the_AT idea_NN1 to_TO have_VHI BLSTR_NP1 work_NN1 in_II parallel_NN1 with_IW RSTP/MSTP_NP1 (_( Section_NN1 ?_? 3_MC )_) ._. 
</s>
<s>
The_AT best_JJT point_NN1 found_VVN is_VBZ then_RT diverted_VVN to_II a_AT1 simplex_NN1 search_NN1 ._. 
</s>
<s>
One_PN1 can_VM employ_VVI parallel_RR direct_JJ solvers_NN2 &lsqb;_( 27_MC &rsqb;_) and_CC &lsqb;_( 28_MC &rsqb;_) or_CC parallel_RR iterative_JJ methods_NN2 ._. 
</s>
<s>
The_AT results_NN2 of_IO the_AT systematic_JJ mapping_NN1 study_NN1 were_VBDR considered_VVN with_II31 respect_II32 to_II33 approaches_NN2 in_II the_AT software_NN1 engineering_NN1 domain_NN1 ._. 
</s>
<s>
Furthermore_RR ,_, this_DD1 example_NN1 is_VBZ by_RR31 no_RR32 means_RR33 an_AT1 isolated_JJ case_NN1 ._. 
</s>
<s>
The_AT conflict_NN1 matrixMcaptures_VVZ the_AT conflict_NN1 among_II tasks_NN2 based_VVN on_II the_AT set_NN1 of_IO tasks_NN2 and_CC the_AT computational_JJ and_CC communicational_JJ conflict_NN1 factors_NN2 ._. 
</s>
<s>
There_EX is_VBZ also_RR the_AT attitude_NN1 that_CST inter-organizational_JJ communication_NN1 via_II hubs_NN2 is_VBZ a_AT1 backward_JJ step_NN1 :_: '_NULL it_PPH1 '_NULL s_ZZ1 password_NN1 this_DD1 ,_, password_NN1 that_CST ,_, its_APPGE just_RR a_AT1 minefield_NN1 of_IO repetitive_JJ actions_NN2 '_NULL (_( Purchasing_VVG Manager_NN1 )_) ._. 
</s>
<s>
This_DD1 shows_VVZ that_CST the_AT algorithm_NN1 is_VBZ more_RRR influenced_VVN by_II the_AT average_JJ costs_NN2 ,_, rather_II21 than_II22 the_AT absolute_JJ value_NN1 ._. 
</s>
<s>
Sequence_NN1 of_IO system_NN1 calls_NN2 produced_VVN by_II integer_NN1 overflow_NN1 exploit_NN1 includes_VVZ over_RG 39,000_MC calls_NN2 ._. 
</s>
<s>
The_AT participating_JJ subjects_NN2 provided_VVN feedback_NN1 regarding_II the_AT usefulness_NN1 of_IO the_AT NLtoSTD-BB_NN1 V2.0_FO method_NN1 and_CC the_AT fault-checklist_JJ method_NN1 ._. 
</s>
<s>
The_AT current_JJ implementation_NN1 supports_VVZ authentication_NN1 to_II a_AT1 local_JJ username-password_JJ database_NN1 according_II21 to_II22 the_AT Open_JJ Web_NN1 Application_NN1 Security_NN1 Project_NN1 ._. 
</s>
<s>
To_TO complete_VVI the_AT proof_NN1 of_IO our_APPGE main_JJ result_NN1 ,_, note_VV0 that_CST ,_, from_II the_AT definition(4)of_FO ,_, we_PPIS2 have_VH0 ._. 
</s>
<s>
This_DD1 growth_NN1 phenomena_NN2 is_VBZ not_XX unique_JJ to_II biometrics_NN2 and_CC has_VHZ been_VBN replicated_VVN in_II many_DA2 other_JJ systems_NN2 which_DDQ seek_VV0 to_TO safeguard_VVI information_NN1 and_CC money_NN1 ._. 
</s>
<s>
Since_CS primitive_JJ propositions_NN2 are_VBR interpreted_VVN relative_II21 to_II22 players_NN2 ,_, we_PPIS2 must_VM allow_VVI the_AT interpretation_NN1 of_IO arbitrary_JJ formulas_NN2 to_TO depend_VVI on_II the_AT player_NN1 as_RR21 well_RR22 ._. 
</s>
<s>
The_AT corresponding_JJ Morven_JJ model_NN1 is_VBZ shown_VVN in_II Table_NN1 3_MC ._. 
</s>
<s>
Spatial_JJ counterparts_NN2 of_IO integral_JJ activity_NN1 are_VBR shown_VVN in_II Fig._NN1 4_MC ._. 
</s>
<s>
Videoconferencing_NN1 software_NN1 typically_RR allows_VVZ users_NN2 to_TO select_VVI codecs_NN2 ._. 
</s>
<s>
Thus_RR ,_, we_PPIS2 may_VM form_VVI our_APPGE entire_JJ equation_NN1 for_IF g_ZZ1 as_II a_AT1 function_NN1 of_IO δ_FO θ_NULL b_ZZ1 ._. 
</s>
<s>
The_AT succeeding_JJ join_NN1 operation_NN1 will_VM join_VVI on_II variable_NN1 v2_FO ._. 
</s>
<s>
Given_VVN the_AT current_JJ level_NN1 of_IO scientific_JJ knowledge_NN1 in_II this_DD1 field_NN1 ,_, a_AT1 modicum_NN1 of_IO uncertainty_NN1 is_VBZ acceptable_JJ ._. 
</s>
<s>
Other_JJ security_NN1 activities_NN2 include_VV0 updating_VVG applications_NN2 ,_, installing_VVG patches_NN2 ,_, turning_VVG off_RP unnecessary_JJ ports_NN2 ,_, and_CC configuring_VVG firewalls_NN2 (_( Rosenthal_NP1 ,_, 2002_MC ,_, Stanton_NP1 et_RA21 al._RA22 ,_, 2003_MC and_CC Whitman_NP1 ,_, 2003_MC )_) ._. 
</s>
<s>
It_PPH1 can_VM be_VBI easily_RR verified_VVN that_CST their_APPGE worst-case_JJ performance_NN1 in_II relation_NN1 with_IW the_AT optimal_JJ solution_NN1 can_VM be_VBI very_RG unsatisfactory_JJ (_( (_( n_ZZ1 )_) guards_VVZ when_RRQ O(1)_FO are_VBR sufficient_JJ )_) ._. 
</s>
<s>
Thus_RR ,_, most_DAT Mutants_NN2 were_VBDR stored_VVN as_CSA simple_JJ records_NN2 that_CST had_VHD a_AT1 replacement_NN1 MIL_NN1 statement_NN1 and_CC an_AT1 index_NN1 into_II the_AT MIL_NN1 ._. 
</s>
<s>
We_PPIS2 identify_VV0 two_MC dimensions_NN2 to_II this_DD1 class_NN1 of_IO problems_NN2 :_: the_AT management_NN1 of_IO background_NN1 events_NN2 and_CC their_APPGE interaction_NN1 with_IW trajectory_NN1 constraints_NN2 ,_, and_CC the_AT management_NN1 of_IO global_JJ propagation_NN1 of_IO effects_NN2 ._. 
</s>
<s>
The_AT other_JJ two_MC features_NN2 represent_VV0 a_AT1 global_JJ point_NN1 of_IO view_NN1 about_II the_AT writing_NN1 in_II the_AT frame_NN1 from_II which_DDQ they_PPHS2 are_VBR extracted_VVN ._. 
</s>
<s>
Subsequently_RR ,_, we_PPIS2 rebuild_VV0 the_AT graph_NN1 without_IW considering_VVG this_DD1 FQDN_NP1 ._. 
</s>
<s>
Rather_RR ,_, we_PPIS2 argue_VV0 that_DD1 given_VVN the_AT prevalence_NN1 of_IO the_AT social_JJ constructs_NN2 within_II these_DD2 sites_NN2 ,_, that_DD1 value_NN1 of_IO the_AT network_NN1 effect_NN1 is_VBZ coming_VVG from_II the_AT links_NN2 between_II people_NN arising_VVG from_II the_AT interactions_NN2 using_VVG these_DD2 sites_NN2 ._. 
</s>
<s>
Log–log_VV0 plots_NN2 show_VV0 (_( a_ZZ1 )_) displacement_NN1 error_NN1 uu_NNU and_CC (_( b_ZZ1 )_) strain_VV0 error_NN1 EE_UH versus_II element_NN1 diameter_NN1 h_ZZ1 ._. 
</s>
<s>
Fundamental_JJ to_II this_DD1 approach_NN1 is_VBZ the_AT notion_NN1 of_IO travel_NN1 time_NNT1 ._. 
</s>
<s>
Although_CS there_EX is_VBZ a_AT1 predefined_JJ set_NN1 of_IO resolution_NN1 policies_NN2 ,_, these_DD2 may_VM be_VBI varied_VVN according_II21 to_II22 the_AT needs_NN2 of_IO the_AT organisation_NN1 ._. 
</s>
<s>
There_EX have_VH0 been_VBN few_DA2 studies_NN2 that_CST consider_VV0 defects_NN2 in_II software_NN1 product_NN1 lines_NN2 or_CC mine_NN1 their_APPGE bug/change_NN1 tracking_NN1 databases_NN2 ._. 
</s>
<s>
However_RR ,_, there_EX is_VBZ no_AT real_JJ guarantee_NN1 that_CST these_DD2 are_VBR the_AT right/appropriate_JJ characteristics_NN2 of_IO real_JJ data_NN ,_, which_DDQ again_RT raises_VVZ the_AT question_NN1 of_IO representativeness_NN1 and_CC realism_NN1 ._. 
</s>
<s>
By_II replacing_VVG the_AT micro-topography_NN1 with_IW spatially_RR distributed_VVN storage_NN1 zones_NN2 ,_, subsurface_NN1 residence_NN1 times_NNT2 remain_VV0 power_NN1 law_NN1 distributed_VVN ,_, as_CSA indicated_VVN by_II the_AT good_JJ fit_NN1 of_IO the_AT p-rs-high_JJ model_NN1 (_( R2_FO ?_? =_FO ?_? 0.9002_MC )_) ._. 
</s>
<s>
Each_DD1 run_NN1 includes_VVZ a_AT1 different_JJ scenario_NN1 of_IO simulated_JJ events_NN2 ,_, applied_VVN on_II the_AT identical_JJ testing_NN1 data_NN set_VV0 ._. 
</s>
<s>
The_AT lower_JJR level_JJ problem_NN1 ,_, however_RR ,_, is_VBZ a_AT1 MIP_NN1 ,_, and_CC there_RL exist_VV0 no_AT close_JJ form_NN1 optimality_NN1 conditions_NN2 for_IF MIPs_NNU (_( except_CS in_II some_DD special_JJ cases_NN2 )_) ._. 
</s>
<s>
Four_MC identical_JJ workshops_NN2 were_VBDR run_VVN within_II the_AT conference_NN1 with_IW 20_MC participants_NN2 in_II each_DD1 ,_, resulting_VVG in_II a_AT1 total_NN1 of_IO 80_MC participants_NN2 ._. 
</s>
<s>
The_AT sender_NN1 '_NULL s_ZZ1 UA_NP1 submits_VVZ a_AT1 message_NN1 (_( either_RR an_AT1 XML_NP1 document_NN1 or_CC binary_JJ content_NN1 )_) to_II the_AT MTA_NN1 of_IO her_APPGE ERV_NN1 provider_NN1 ._. 
</s>
<s>
A_AT1 one-size-fits-all_DB solution_NN1 does_VDZ not_XX exist_VVI and_CC would_VM not_XX be_VBI appropriate_JJ given_VVN the_AT wide_JJ variety_NN1 of_IO business_NN1 missions_NN2 ,_, criticality_NN1 of_IO information_NN1 ,_, and_CC BYOD_VV0 implementations_NN2 that_CST may_VM exist_VVI from_II organization_NN1 to_II organization_NN1 ._. 
</s>
<s>
For_IF larger_JJR polygons_NN2 ,_, the_AT search_NN1 may_VM initially_RR only_RR treat_VVI prime_JJ numbers_NN2 of_IO sides_NN2 ,_, and_CC explore_VV0 further_RRR if_CS these_DD2 have_VH0 a_AT1 high_JJ likelihood_NN1 ._. 
</s>
<s>
Therefore_RR ,_, the_AT database_NN1 includes_VVZ the_AT effect_NN1 of_IO age_NN1 variations_NN2 ._. 
</s>
<s>
We_PPIS2 can_VM now_RT formally_RR define_VVI the_AT different_JJ manipulation_NN1 problems_NN2 we_PPIS2 consider_VV0 ._. 
</s>
<s>
The_AT above_JJ Corollary_NN1 1_MC1 and_CC Theorem_NN1 11_MC indicate_VV0 that_CST in_II certain_JJ restricted_JJ domains_NN2 ,_, verifying_VVG -core_JJ imputations_NN2 (_( or_CC computing_VVG the_AT maximal_JJ excess_NN1 under_II a_AT1 given_JJ imputation_NN1 )_) can_VM be_VBI performed_VVN in_II polynomial_NN1 time_NNT1 ._. 
</s>
<s>
Because_CS biases_NN2 may_VM occur_VVI when_RRQ formative_JJ constructs_NN2 are_VBR mis-specified_VVN as_CSA reflective_JJ (_( Mackenzie_NP1 et_RA21 al._RA22 ,_, 2005_MC )_) ,_, the_AT constructs_NN2 were_VBDR classified_VVN according_II21 to_II22 the_AT four_MC decision_NN1 rules_NN2 outlined_VVN by_II (_( Petter_NP1 et_RA21 al._RA22 ,_, 2007_MC )_) ._. 
</s>
<s>
Being_VBG the_AT final_JJ formula_NN1 f_ZZ1 a_AT1 sequence_NN1 of_IO Tf_NN1 symbols_NN2 ,_, the_AT decoder_NN1 will_VM produce_VVI one_MC1 activation_NN1 vector_NN1 for_IF each_DD1 of_IO these_DD2 symbols_NN2 ._. 
</s>
<s>
Imagine_VV0 that_CST a_AT1 user_NN1 submits_VVZ a_AT1 query_NN1 to_II peerA_NN1 ._. 
</s>
<s>
PeerAwill_VV0 first_MD process_VV0 the_AT query_NN1 and_CC return_VV0 any_DD search_NN1 results_VVZ it_PPH1 finds_VVZ to_II the_AT user_NN1 ._. 
</s>
<s>
This_DD1 means_VVZ that_CST the_AT proposed_JJ algorithm_NN1 becomes_VVZ more_RGR efficient_JJ than_CSN the_AT GKM_NP1 and_CC MGKM_NP1 algorithms_NN2 when_RRQ the_AT size_NN1 of_IO the_AT datasets_NN2 increases_VVZ ._. 
</s>
<s>
They_PPHS2 can_VM not_XX ,_, however_RR ,_, quantify_VV0 over_RP situation_NN1 terms_NN2 or_CC compare_VV0 situations_NN2 using_VVG a_AT1 ≤_FO relation_NN1 ._. 
</s>
<s>
In_BCL21 order_BCL22 to_TO reduce_VVI the_AT computational_JJ cost_NN1 ,_, integral_JJ image_NN1 representation_NN1 &lsqb;_( 39_MC &rsqb;_) ,_, &lsqb;_( 6_MC &rsqb;_) is_VBZ used_VVN to_TO obtain_VVI the_AT SWS_NP1 ._. 
</s>
<s>
A_AT1 researcher_NN1 '_NULL s_ZZ1 first-hand_JJ experience_NN1 with_IW different_JJ research_NN1 sites_NN2 is_VBZ essential_JJ for_IF capturing_VVG contextual_JJ knowledge_NN1 of_IO locations_NN2 and_CC to_TO build_VVI a_AT1 knowledge_NN1 base_NN1 that_CST can_VM be_VBI used_VVN to_TO interpret_VVI the_AT data_NN in_II a_AT1 meaningfully_RR ._. 
</s>
<s>
For_REX21 example_REX22 ,_, Arc_NP1 Hydro_NP1 data_NN model_NN1 integrates_VVZ the_AT monitoring_NN1 locations_NN2 and_CC their_APPGE time_NNT1 series_NN data_NN within_II the_AT GIS_NN2 database_NN1 ._. 
</s>
<s>
This_DD1 unit_NN1 needs_VVZ to_TO examine_VVI the_AT header_NN1 of_IO each_DD1 packet_NN1 which_DDQ is_VBZ being_VBG added_VVN or_CC dropped_VVN at_II that_DD1 node_NN1 and_CC selectively_RR duplicate_VV0 the_AT packets_NN2 ._. 
</s>
<s>
In_II practice_NN1 ,_, because_CS they_PPHS2 require_VV0 an_AT1 additional_JJ relationship_NN1 type_NN1 and_CC entity_NN1 type_NN1 to_TO be_VBI included_VVN ,_, these_DD2 are_VBR often_RR left_VVN off_II conceptual_JJ diagrams_NN2 (_( seeFig._NNU 1_MC1 )_) with_IW their_APPGE existence_NN1 noted_VVN elsewhere.2More_FO complex_JJ domains_NN2 ,_, such_II21 as_II22 hierarchies_NN2 ,_, can_VM not_XX be_VBI easily_RR included_VVN at_RR21 all_RR22 ._. 
</s>
<s>
We_PPIS2 are_VBR not_XX claiming_VVG that_CST the_AT Hardware_NN1 Trojan_JJ that_CST we_PPIS2 have_VH0 developed_VVN would_VM evade_VVI these_DD2 detection_NN1 techniques_NN2 ._. 
</s>
<s>
E-collaboration_NN1 was_VBDZ particularly_RR important_JJ for_IF our_APPGE Agile_JJ teams_NN2 because_CS (_( a_ZZ1 )_) Agile_JJ requires_VVZ regular_JJ customer_NN1 involvement_NN1 (_( b_ZZ1 )_) several_DA2 teams_NN2 had_VHD physically_RR distant_JJ customers_NN2 making_VVG face-to-face_JJ collaboration_NN1 difficult_JJ and_CC (_( c_ZZ1 )_) e-collaboration_NN1 provided_VVD cheaper_JJR alternative_NN1 ._. 
</s>
<s>
TTP_NP1 :_: the_AT average_JJ interval_NN1 between_II when_RRQ a_AT1 target_NN1 is_VBZ first_MD aware_JJ of_IO the_AT existence_NN1 of_IO a_AT1 new_JJ threat_NN1 and_CC when_CS it_PPH1 successfully_RR deflects_VVZ it_PPH1 ._. 
</s>
<s>
The_AT averaging_NN1 scheme_NN1 is_VBZ not_XX the_AT best_JJT choice_NN1 ,_, but_CCB is_VBZ simple_JJ and_CC easy_JJ to_TO implement_VVI ._. 
</s>
<s>
The_AT gravitational_JJ force_NN1 acts_VVZ downwards_RL with_IW @g=9.81m/s2_FU and_CC air_NN1 is_VBZ neglected_VVN in_II simulations_NN2 ._. 
</s>
<s>
Routers_NP1 B_ZZ1 and_CC C_ZZ1 perform_VV0 RPF_NP1 check_NN1 and_CC the_AT check_NN1 succeeds_VVZ for_IF both_DB2 routers_NN2 ._. 
</s>
<s>
During_II the_AT addition_NN1 of_IO new_JJ reviewers_NN2 to_II the_AT document_NN1 ,_, due_JJ to_TO conflict_VVI of_IO interest_NN1 reasons_NN2 ,_, a_AT1 second_MD transaction_NN1 must_VM not_XX insert_VVI new_JJ authors_NN2 to_II the_AT document_NN1 ,_, because_CS the_AT determination_NN1 of_IO conflict_NN1 for_IF the_AT document_NN1 is_VBZ based_VVN on_II its_APPGE current_JJ authors_NN2 ._. 
</s>
<s>
However_RR ,_, even_CS21 if_CS22 the_AT attacker_NN1 was_VBDZ to_TO successfully_RR impersonate_VVI a_AT1 valid_JJ tag_NN1 ,_, on_II reception_NN1 of_IO ?_? 
</s>
<s>
Gi_NN1 ,_, the_AT attacker_NN1 will_VM be_VBI required_VVN to_II firstly_RR ,_, extract_VV0 ?_? 
</s>
<s>
ACKi_NN2 ?_? from_II ?_? 
</s>
<s>
Gi_NN1 ._. 
</s>
<s>
The_AT former_DA case_NN1 should_VM improve_VVI his_APPGE reputation_NN1 while_CS the_AT latter_DA one_PN1 should_VM reduce_VVI it_PPH1 since_CS he_PPHS1 is_VBZ requesting_VVG more_RGR often_RR than_CSN other_JJ cooperative_JJ participants_NN2 ._. 
</s>
<s>
Since_CS itemsd_NN1 ,_, f_ZZ1 ,_, andmdo_NN1 not_XX meet_VV0 the_AT minimum_JJ frequency_NN1 of_IO 3_MC ,_, they_PPHS2 are_VBR not_XX inserted_VVN inn_NN1 '_NULL s_ZZ1 conditional_JJ FSP.Tree_NP1 ._. 
</s>
<s>
The_AT success_NN1 evaluating_VVG technique_NN1 chosen_VVN in_II the_AT strategy_NN1 stage_NN1 is_VBZ applied_VVN here_RL ._. 
</s>
<s>
These_DD2 can_VM include_VVI token_JJ codes_NN2 ,_, common_JJ access_NN1 cards_NN2 ,_, or_CC biometrics_NN2 ._. 
</s>
<s>
Moreover_RR ,_, PDDL_NP1 supports_VVZ optimization_NN1 so_CS21 that_CS22 the_AT specific_JJ metrics_NN2 (_( i.e._REX ,_, NFPs_NP1 in_II our_APPGE context_NN1 )_) can_VM be_VBI minimized_VVN or_CC maximized_VVD during_II plan_NN1 generation_NN1 ._. 
</s>
<s>
An_AT1 SBS_NP1 scheme_NN1 in_II a_AT1 router_NN1 or_CC server_NN1 gives_VVZ priority_NN1 to_II flows_NN2 depending_II21 on_II22 their_APPGE size_NN1 ._. 
</s>
<s>
However_RR ,_, last_MD few_DA2 years_NNT2 ,_, buffer_NN1 sizing_NN1 of_IO router_NN1 attracted_VVD lots_PN of_IO attention_NN1 ._. 
</s>
<s>
Content_JJ validity_NN1 for_IF all_DB instrument_NN1 scales_NN2 was_VBDZ established_VVN through_II both_DB2 literature_NN1 review_NN1 and_CC an_AT1 expert_NN1 panel_NN1 comprised_VVN of_IO 3_MC researchers_NN2 with_IW experience_NN1 in_II scale_NN1 development_NN1 and_CC 2_MC information_NN1 security_NN1 experts_NN2 ._. 
</s>
<s>
We_PPIS2 illustrate_VV0 in_II the_AT following_JJ how_RRQ the_AT resilience_NN1 of_IO this_DD1 approach_NN1 can_VM be_VBI improved_VVN by_II introducing_VVG redundancy_NN1 in_II storing_VVG the_AT original_JJ data_NN ._. 
</s>
<s>
In_II the_AT following_JJ figures_NN2 ,_, all_DB data_NN points_NN2 shown_VVN are_VBR averages_NN2 over_RG 100_MC simulation_NN1 runs_NN2 in_BCL21 order_BCL22 to_TO capture_VVI the_AT randomized_JJ nature_NN1 of_IO the_AT algorithms_NN2 investigated_VVD ._. 
</s>
<s>
We_PPIS2 will_VM explore_VVI two_MC estimators_NN2 that_CST provide_VV0 useful_JJ representations_NN2 of_IO the_AT probability_NN1 distributions_NN2 underlying_VVG such_DA data_NN ._. 
</s>
<s>
In_II our_APPGE work_NN1 ,_, we_PPIS2 have_VH0 assumed_VVN VD_NN1 requests_NN2 can_VM arise_VVI from_II three_MC kinds_NN2 of_IO user_NN1 profiles_NN2 -_- a_AT1 student_NN1 in_II campus_NN1 computer_NN1 lab_NN1 ,_, a_AT1 distance_NN1 learning_NN1 user_NN1 or_CC a_AT1 scientist_NN1 in_II Engineering_NN1 Site_NN1 ,_, each_DD1 having_VHG specific_JJ resource_NN1 requirements_NN2 needed_VVN for_IF satisfactory_JJ QoE_NN1 of_IO their_APPGE VD_NN1 application_NN1 sets_NN2 ._. 
</s>
<s>
We_PPIS2 implemented_VVD the_AT SemTag_NN1 algorithm_NN1 described_VVN earlier_RRR ,_, and_CC applied_VVN it_PPH1 to_II a_AT1 set_NN1 of_IO 264_MC million_NNO pages_NN2 producing_VVG 270G_NNU of_IO dump_NN1 data_NN corresponding_VVG to_II 550_MC million_NNO labels_NN2 in_II context_NN1 ._. 
</s>
<s>
The_AT chess_NN1 engine_NN1 was_VBDZ run_VVN in_II MultiPV_NP1 (_( i.e._REX ,_, multi-best-line_NN1 )_) mode_NN1 and_CC set_VV0 to_TO return_VVI evaluations_NN2 for_IF the_AT five_MC best_JJT moves_NN2 in_II each_DD1 position_NN1 (_( but_CCB the_AT analyser_NN1 allows_VVZ the_AT number_NN1 of_IO moves_NN2 to_TO be_VBI varied_VVN as_CSA required_VVN )_) ._. 
</s>
<s>
Using_VVG (_( 2.8_MC )_) with_IW the_AT normalization_NN1 (_( 1.2_MC )_) we_PPIS2 can_VM expand_VVI k(uR)_NN1 as_CSA ,_, Error-correcting_JJ graph_NN1 techniques_NN2 make_VV0 the_AT algorithm_NN1 resilient_JJ against_II edge-flip_JJ attacks_NN2 ,_, in_II which_DDQ the_AT basic_JJ blocks_NN2 are_VBR reordered_VVN ,_, but_CCB it_PPH1 remains_VVZ vulnerable_JJ to_II a_AT1 large_JJ number_NN1 of_IO other_JJ semantics-preserving_JJ code_NN1 transformations_NN2 ._. 
</s>
<s>
As_CSA in_II PA_NN1 ,_, qualitative_JJ constraints_NN2 in_II a_AT1 Morven_JJ model_NN1 are_VBR distributed_VVN over_II multiple_JJ differential_JJ planes_NN2 ._. 
</s>
<s>
The_AT first_MD several_DA2 eigenvectors_NN2 capture_VV0 the_AT relative_JJ placement_NN1 of_IO these_DD2 features_NN2 (_( Fig._NN1 10a_FO )_) ._. 
</s>
<s>
However_RRQV this_DD1 assumption_NN1 is_VBZ a_AT1 substantive_JJ one_PN1 ._. 
</s>
<s>
This_DD1 is_VBZ caused_VVN by_II the_AT large_JJ amount_NN1 of_IO versions_NN2 that_CST are_VBR stored_VVN for_IF each_DD1 tree_NN1 value_NN1 ._. 
</s>
<s>
NAST_NP1 (_( Section4.3.5_FO )_) with_IW ann_NN1 -gram_JJ length_NN1 ofn=2_FO ._. 
</s>
<s>
Process_VV0 and_CC user_NN1 interface_NN1 :_: Another_DD1 issue_NN1 that_CST Mothra_NP1 helped_VVD consider_VVI was_VBDZ how_RRQ best_JJT to_II structure_NN1 the_AT mutation_NN1 process_NN1 ._. 
</s>
<s>
This_DD1 was_VBDZ a_AT1 complex_JJ project_NN1 in_II31 view_II32 of_II33 its_APPGE large_JJ cost_NN1 and_CC scope_NN1 ,_, significant_JJ level_NN1 of_IO integration_NN1 ,_, and_CC technical_JJ and_CC task_NN1 complexity_NN1 ._. 
</s>
<s>
Note_VV0 that_CST full_JJ connectivity_NN1 in_II the_AT presence_NN1 of_IO continuous_JJ churn_NN1 is_VBZ a_AT1 desired_JJ property_NN1 of_IO any_DD routing_NN1 infrastructure_NN1 ._. 
</s>
<s>
For_REX21 instance_REX22 ,_, if_CS the_AT location_NN1 is_VBZ a_AT1 sphere_NN1 we_PPIS2 store_VV0 the_AT center_NN1 of_IO the_AT sphere_NN1 and_CC the_AT radius_NN1 ._. 
</s>
<s>
How_RRQ can_VM I_PPIS1 sample_VVI a_AT1 real_JJ networkWe_NN1 conclude_VV0 from_II our_APPGE experiments_NN2 that_CST DHYB-0.8_MC is_VBZ the_AT best_RRT among_II our_APPGE methods_NN2 for_IF the_AT Internet_NN1 sampling_NN1 ,_, and_CC that_CST it_PPH1 also_RR compares_VVZ favorably_RR to_II graph_NN1 generation_NN1 methods_NN2 proposed_VVD previously_RR in_II the_AT literature_NN1 ._. 
</s>
<s>
For_REX21 example_REX22 ,_, the_AT efficacy_NN1 of_IO a_AT1 drug_NN1 can_VM be_VBI framed_VVN in_II31 terms_II32 of_II33 number_NN1 of_IO lives_NN2 saved_VVN or_CC number_NN1 of_IO lives_NN2 lost_VVN ;_; studies_NN2 have_VH0 shown_VVN that_CST equivalent_JJ data_NN framed_VVN in_II opposite_JJ ways_NN2 (_( gain_VV0 vs._II loss_NN1 )_) lead_VV0 to_TO dramatically_RR different_JJ decisions_NN2 about_II whether_CSW and_CC how_RRQ to_TO use_VVI the_AT same_DA drug_NN1 ._. 
</s>
<s>
Driven_JJ (_( in_RR21 part_RR22 )_) by_II the_AT success_NN1 and_CC achievements_NN2 of_IO TREC_NP1 ,_, a_AT1 number_NN1 of_IO other_JJ large-scale_JJ IR_NN1 evaluations_NN2 have_VH0 been_VBN run_VVN ._. 
</s>
<s>
Furthermore_RR ,_, our_APPGE software_NN1 made_VVD the_AT service_NN1 recommendations_NN2 within_II minutes_NNT2 while_CS it_PPH1 took_VVD human_JJ architects_NN2 about_RG 9_MC h_ZZ1 to_TO develop_VVI their_APPGE suggestions_NN2 and_CC then_RT several_DA2 additional_JJ months_NNT2 to_TO refine_VVI their_APPGE suggestions_NN2 ._. 
</s>
<s>
The_AT respective_JJ size_NN1 of_IO the_AT PostGIS_NN1 database_NN1 that_CST was_VBDZ produced_VVN after_II loading_VVG the_AT RDF_NP1 dump_NN1 to_II Strabon_NP1 is_VBZ 1665_MC MB_NNU ,_, which_DDQ is_VBZ more_DAR than_CSN twice_RR the_AT disk_NN1 space_NN1 compared_VVN to_II the_AT original_JJ database_NN1 produced_VVN by_II importing_VVG the_AT shapefiles_NN2 directly_RR ._. 
</s>
<s>
In_II Fig._NN1 4A_FO ,_, we_PPIS2 give_VV0 the_AT specification_NN1 of_IO the_AT template_NN1 entity_NN1 "_" Water_NN1 "_" ,_, which_DDQ can_VM be_VBI described_VVN by_II a_AT1 single_JJ variable_NN1 "_" amountOfWater_NN1 "_" ._. 
</s>
<s>
The_AT Audit_NN1 was_VBDZ 32_MC pages_NN2 long_JJ &lsqb;_( DM1_FO ,_, Table_NN1 1_MC1 &rsqb;_) ,_, organised_VVN in_II a_AT1 series_NN of_IO short_JJ chapters_NN2 ,_, each_DD1 addressing_VVG a_AT1 particular_JJ theme_NN1 :_: youth_NN1 crime_NN1 ,_, domestic_JJ burglary_NN1 ,_, etc_RA ._. 
</s>
<s>
Phase_NN1 4:The_FO attacker_NN1 uses_VVZ telnet_NN1 and_CC rpc_NNU to_TO install_VVI a_AT1 DDoS_NP1 program_NN1 in_II the_AT compromised_JJ machines_NN2 ._. 
</s>
<s>
Our_APPGE proposed_JJ scheme_NN1 ensures_VVZ the_AT flexibility_NN1 since_CS it_PPH1 is_VBZ able_JK to_TO extend_VVI with_IW the_AT increase/decrease_NN1 of_IO the_AT number_NN1 of_IO packets_NN2 in_II generation_NN1 ?_? h_ZZ1 ?_? ._. 
</s>
<s>
Using_VVG the_AT bidomain_NN1 model_NN1 &lsqb;_( 3_MC &rsqb;_) ,_, more_DAR than_CSN two_MC CPU_NN1 weeks_NNT2 are_VBR required_VVN to_TO simulate_VVI 400_MC ms_NN1 of_IO cardiac_JJ activity_NN1 on_II the_AT rabbit_NN1 mesh_NN1 using_VVG the_AT Chaste_JJ software_NN1 package_NN1 &lsqb;_( 4_MC &rsqb;_) ._. 
</s>
<s>
This_DD1 analysis_NN1 would_VM suggest_VVI further_JJR explorations_NN2 into_II modifications_NN2 of_IO other_JJ efficient_JJ methods_NN2 ._. 
</s>
<s>
One_MC1 of_IO such_DA visualizations_NN2 can_VM be_VBI seen_VVN in_II Fig._NN1 10_MC ._. 
</s>
<s>
They_PPHS2 come_VV0 with_IW the_AT understanding_NN1 that_CST this_DD1 is_VBZ a_AT1 temporary_JJ contract_NN1 ._. 
</s>
<s>
We_PPIS2 rely_VV0 primarily_RR on_II total_JJ degrees_NN2 of_IO freedom_NN1 here_RL ,_, since_CS the_AT majority_NN1 of_IO the_AT computations_NN2 for_IF the_AT methods_NN2 compared_VVN are_VBR implemented_VVN within_II the_AT same_DA general-purpose_JJ finite_JJ element_NN1 library_NN1 ._. 
</s>
<s>
Table_NN1 4_MC shows_VVZ the_AT results_NN2 with_II31 respect_II32 to_II33 the_AT type_NN1 of_IO the_AT answer_NN1 that_CST the_AT questions_NN2 expect_VV0 ,_, independent_JJ of_IO whether_CSW they_PPHS2 require_VV0 aggregations_NN2 or_CC another_DD1 schema_NN1 ,_, comprising_VVG the_AT following_JJ answer_NN1 types_NN2 :_: numbers_NN2 or_CC a_AT1 count_NN1 of_IO the_AT result_NN1 answers_NN2 (_( 15_MC queries_NN2 )_) ,_, literals_NN2 or_CC strings_NN2 (_( 3_MC queries_NN2 )_) ,_, booleans_NN2 (_( 8_MC queries_NN2 )_) ,_, dates_NN2 (_( 3_MC queries_NN2 )_) and_CC a_AT1 resource_NN1 or_CC a_AT1 list_NN1 of_IO resources_NN2 (_( 70_MC queries_NN2 )_) ._. 
</s>
<s>
This_DD1 research_NN1 began_VVD when_CS we_PPIS2 noticed_VVD some_DD of_IO the_AT weakness_NN1 of_IO client-side_JJ PKI_NN2 and_CC browser_NN1 keystores_VVZ in_II the_AT literature_NN1 ._. 
</s>
<s>
Using_VVG the_AT solved_JJ labels_NN2 ,_, the_AT sampling_NN1 procedure_NN1 of_IO LRTDP_NP1 can_VM be_VBI seen_VVN as_II a_AT1 case_NN1 of_IO rejection_NN1 sampling_NN1 :_: if_CS the_AT sampled_JJ successor_NN1 s_ZZ1 of_IO s_ZZ1 is_VBZ marked_VVN as_CSA solved_VVN ,_, restart_VV0 the_AT procedure_NN1 from_II the_AT initial_JJ state_NN1 s0_FO ;_; otherwise_RR ,_, use_VV0 s_ZZ1 ._. 
</s>
<s>
This_DD1 is_VBZ a_AT1 process-based_JJ model_NN1 (_( Landsberg_NP1 and_CC Waring_NP1 ,_, 1997_MC )_) commonly_RR used_VMK to_TO predict_VVI tree_NN1 and_CC forest_NN1 growth_NN1 for_IF a_AT1 range_NN1 of_IO species_NN (_( e.g._REX Almeida_NP1 et_NP1 ?_? al._NNU ,_, 2004_MC ,_, Paul_NP1 et_NP1 ?_? al._NNU ,_, 2008_MC )_) under_II different_JJ climate_NN1 ,_, soil_NN1 ,_, and_CC management_NN1 conditions_NN2 (_( Crossman_NP1 et_NP1 ?_? al._NNU ,_, 2011a_FO ,_, Paterson_NP1 and_CC Bryan_NP1 ,_, 2012_MC ,_, Paul_NP1 et_NP1 ?_? al._NNU ,_, 2013a_FO ,_, Paul_NP1 et_NP1 ?_? al._NNU ,_, 2013b_FO )_) ._. 
</s>
<s>
For_IF computing_VVG the_AT average_NN1 ,_, all_DB weights_NN2 have_VH0 to_TO be_VBI initialized_VVN identically_RR towi=1_FO ._. 
</s>
<s>
Nevertheless_RR ,_, the_AT approximability_NN1 of_IO isogeometric_JJ analysis_NN1 compared_VVN with_IW classical_JJ finite_JJ element_NN1 analysis_NN1 has_VHZ not_XX been_VBN thoroughly_RR investigated_VVN ._. 
</s>
<s>
Learning_NN1 involves_VVZ relating_VVG parts_NN2 of_IO the_AT subject_NN1 matter_NN1 to_II each_PPX221 other_PPX222 and_CC to_II the_AT real_JJ world_NN1 ._. 
</s>
<s>
One_MC1 objective_NN1 consists_VVZ of_IO minimizing_VVG the_AT costs_NN2 of_IO construction_NN1 and_CC operations_NN2 of_IO the_AT network_NN1 ._. 
</s>
<s>
The_AT method_NN1 extends_VVZ our_APPGE earlier_JJR work_NN1 by_II using_VVG a_AT1 4-point_JJ ordinal_JJ scale_NN1 ,_, moderating_VVG the_AT expert_NN1 assessments_NN2 using_VVG argumentation_NN1 theory_NN1 and_CC propagating_VVG the_AT assessments_NN2 using_VVG tabulation_NN1 ._. 
</s>
<s>
An_AT1 action_NN1 moves_VVZ the_AT world_NN1 from_II one_MC1 state_NN1 to_II another_DD1 ,_, where_CS in_II MDPs_NP1 this_DD1 transition_NN1 is_VBZ non-deterministic_JJ ._. 
</s>
<s>
In_RR21 addition_RR22 ,_, optimal_JJ filter_NN1 parameters_NN2 can_VM be_VBI obtained_VVN for_IF catchments_NN2 with_IW different_JJ physical_JJ characteristics_NN2 ,_, which_DDQ will_VM assist_VVI with_IW providing_VVG an_AT1 insight_NN1 into_II the_AT range_NN1 of_IO catchment_NN1 properties_NN2 for_IF which_DDQ different_JJ RDFs_NP1 are_VBR applicable_JJ (_( i.e._REX perform_VV0 adequately_RR )_) ,_, provided_CS the_AT optimal_JJ filter_NN1 parameters_NN2 are_VBR used_VVN (_( referred_VVN to_II as_II the_AT '_NULL range_NN1 of_IO applicability_NN1 '_NULL of_IO different_JJ RDFs_NP1 )_) and_CC the_AT sensitivity_NN1 of_IO optimal_JJ values_NN2 of_IO filter_NN1 parameters_NN2 to_II various_JJ catchment_NN1 properties_NN2 ._. 
</s>
<s>
It_PPH1 is_VBZ important_JJ to_TO note_VVI here_RL that_CST if_CS all_DB or_CC some_DD components_NN2 of_IO a_AT1 software_NN1 application_NN1 are_VBR service_NN1 oriented_VVD abstraction_NN1 specific_JJ then_RT SoaML_NP1 (_( e.g._REX service_NN1 ,_, service_NN1 interface_NN1 )_) can_VM be_VBI used_VVN to_TO model_VVI the_AT service_NN1 oriented_VVD components_NN2 of_IO the_AT software_NN1 application_NN1 in_II detail_NN1 ._. 
</s>
<s>
In_II situations_NN2 where_RRQ perimeter_NN1 defenses_NN2 are_VBR the_AT main_JJ security_NN1 measure_NN1 ,_, this_DD1 allows_VVZ attackers_NN2 access_VV0 to_II sensitive_JJ data_NN ,_, access_NN1 to_II other_JJ internal_JJ machines_NN2 ,_, and_CC can_VM enable_VVI installation_NN1 of_IO backdoor_NN1 programs_NN2 for_IF ongoing_JJ control_NN1 of_IO internal_JJ hosts_NN2 ._. 
</s>
<s>
Note_VV0 that_CST when_CS we_PPIS2 refer_VV0 to_II the_AT "_" order_NN1 "_" of_IO a_AT1 NURBS_NN2 curve_VV0 ,_, we_PPIS2 mean_VV0 the_AT order_NN1 of_IO the_AT polynomial_NN1 curve_NN1 from_II which_DDQ the_AT rational_JJ curve_NN1 was_VBDZ generated_VVN ._. 
</s>
<s>
The_AT rest_NN1 of_IO the_AT paper_NN1 is_VBZ organized_VVN as_CSA follows_VVZ ._. 
</s>
<s>
The_AT primary_JJ drawback_NN1 is_VBZ that_CST a_AT1 piecewise_JJ constant_JJ ,_, average_JJ approximation_NN1 for_IF source_NN1 terms_NN2 and_CC material_NN1 coefficients_NN2 is_VBZ necessary_JJ for_IF the_AT nonconforming_JJ solution_NN1 ._. 
</s>
<s>
For_IF comparison_NN1 ,_, the_AT image_NN1 pair_NN was_VBDZ also_RR registered_VVN using_VVG the_AT affine_JJ model_NN1 for_IF displacement_NN1 field_NN1 (_( d_ZZ1 )_) ._. 
</s>
<s>
In_II this_DD1 manner_NN1 ,_, the_AT survey_NN1 was_VBDZ made_VVN manageable_JJ and_CC easy_JJ to_TO read_VVI ,_, avoiding_VVG excessive_JJ wording_NN1 ._. 
</s>
<s>
Replication_NN1 studies_NN2 are_VBR beneficial_JJ to_TO evaluate_VVI the_AT validity_NN1 of_IO prior_JJ study_NN1 findings_NN2 ,_, either_RR by_II reproducing_VVG results_NN2 or_CC by_II isolating_JJ factors_NN2 that_CST can_VM influence_VVI results_NN2 and_CC that_DD1 lead_NN1 to_II variations_NN2 ._. 
</s>
<s>
The_AT elements_NN2 are_VBR then_RT typically_RR rated_VVN against_II each_DD1 construct_NN1 ._. 
</s>
<s>
However_RR ,_, despite_II the_AT substantial_JJ performance_NN1 gains_NN2 on_II offer_NN1 ,_, the_AT use_NN1 of_IO HPC_NP1 in_II the_AT integrated_JJ assessment_NN1 and_CC modelling_NN1 of_IO social–ecological_JJ systems_NN2 has_VHZ not_XX been_VBN widespread_JJ ._. 
</s>
<s>
Standard_JJ authentication_NN1 is_VBZ based_VVN on_II username_NN1 and_CC password_NN1 ._. 
</s>
<s>
Isogeometric_JJ analysis_NN1 was_VBDZ introduced_VVN by_II Hughes_NP1 et_RA21 al._RA22 &lsqb;_( 15_MC &rsqb;_) in_II an_AT1 effort_NN1 to_TO improve_VVI upon_II shortcomings_NN2 of_IO finite_JJ element_NN1 analysis_NN1 in_II the_AT areas_NN2 of_IO geometric_JJ precision_NN1 ,_, ease_NN1 of_IO mesh_NN1 refinement_NN1 ,_, and_CC integration_NN1 with_IW Computer_NN1 Aided_VVD Design_NN1 (_( CAD_NN1 )_) ._. 
</s>
<s>
The_AT first_MD is_VBZ a_AT1 generalisation_NN1 of_IO an_AT1 algorithm_NN1 presented_VVN in_II &lsqb;_( 25_MC &rsqb;_) and_CC relies_VVZ on_II the_AT credulous_JJ acceptance_NN1 problem_NN1 for_IF ._. 
</s>
<s>
The_AT size_NN1 of_IO the_AT geographical_JJ domain_NN1 used_VVN for_IF the_AT predictors_NN2 (_( latitude_NN1 and_CC longitude_NN1 )_) ._. 
</s>
<s>
This_DD1 behaviour_NN1 is_VBZ observed_VVN on_II the_AT right-hand_JJ sides_NN2 of_IO the_AT graphs_NN2 ._. 
</s>
<s>
Basin_NN1 iv_MC shows_VVZ a_AT1 more_RGR "_" mature_JJ "_" curve_NN1 under_II the_AT modified_JJ condition_NN1 because_CS dams_NN2 capture_VV0 flow_NN1 from_II steeper_JJR parts_NN2 of_IO the_AT upper_JJ catchment_NN1 ._. 
</s>
<s>
It_PPH1 has_VHZ been_VBN shown_VVN that_CST the_AT sum_NN1 rule_NN1 achieves_VVZ the_AT best_JJT classification_NN1 performance_NN1 in_II comparison_NN1 to_II other_JJ measurement-level_JJ fusion_NN1 strategies_NN2 ,_, such_II21 as_II22 product_NN1 and_CC median_JJ rules_NN2 ._. 
</s>
<s>
Weiss_VV0 et_RA21 al_RA22 ._. 
</s>
<s>
(_( 2012_MC )_) presented_VVD the_AT results_NN2 of_IO a_AT1 meta-analysis_NN1 on_II 44_MC LCA_NN1 studies_NN2 on_II biobased_JJ materials_NN2 ,_, and_CC observed_VVD differences_NN2 in_II LCA_NP1 studies_NN2 concerning_II assumptions_NN2 and_CC choices_NN2 in_II system_NN1 boundaries_NN2 ,_, functional_JJ units_NN2 ,_, scenarios_NN2 and_CC allocation_NN1 approaches_NN2 ._. 
</s>
<s>
Both_RR EPIC_JJ and_CC NAST_NP1 are_VBR methods_NN2 for_IF learning_VVG useful_JJ lines_NN2 of_IO play_NN1 for_IF the_AT playout_NN1 policy_NN1 ;_; EPIC_NN1 achieves_VVZ this_DD1 by_II learning_VVG fragments_NN2 of_IO lines_NN2 delimited_VVD by_II game-specific_JJ episodes_NN2 ,_, whereas_CS NAST_NP1 withn=2essentially_FO learns_VVZ a_AT1 Markov_NP1 model_NN1 ._. 
</s>
<s>
But_CCB if_CS the_AT sources_NN2 do_VD0 not_XX provide_VVI consistent_JJ answers_NN2 ,_, a_AT1 not-unlikely_JJ circumstance_NN1 ,_, then_RT a_AT1 person_NN1 has_VHZ complicated_VVN his_APPGE or_CC her_APPGE decision_NN1 making_VVG ._. 
</s>
<s>
We_PPIS2 begin_VV0 our_APPGE experiments_NN2 by_II ranking_VVG the_AT regularity_NN1 metrics_NN2 in_II order_NN1 of_IO highest_JJT individual_JJ classification_NN1 rates_NN2 ._. 
</s>
<s>
Each_DD1 simulation_NN1 was_VBDZ identified_VVN by_II a_AT1 combination_NN1 ofb_NN1 ,_, d_ZZ1 ,_, K_ZZ1 ,_, n_ZZ1 ,_, and_CC W+F_FO values_NN2 ,_, where_CS W+F_FO is_VBZ the_AT total_JJ number_NN1 of_IO join_NN1 and_CC failure_NN1 events_NN2 ._. 
</s>
<s>
An_AT1 incomplete_JJ conceptual_JJ model_NN1 can_VM represent_VVI multiple_JJ conceptualizations_NN2 of_IO the_AT system_NN1 ,_, each_DD1 contributing_VVG to_II the_AT generation_NN1 of_IO at_RR21 least_RR22 one_MC1 candidate_NN1 model_NN1 structure_NN1 (_( Fig._NN1 8_MC )_) ._. 
</s>
<s>
A_AT1 sensitivity_NN1 analysis_NN1 was_VBDZ completed_VVN for_IF the_AT market_NN1 penetration_NN1 rate_NN1 variable_NN1 and_CC is_VBZ presented_VVN in_II Fig._NN1 4_MC ._. 
</s>
<s>
In_II a_AT1 real-world_JJ deployment_NN1 ,_, the_AT number_NN1 of_IO requests_NN2 ?_? may_VM ?_? be_VBI used_VVN as_II an_AT1 additional_JJ feature_NN1 ,_, to_TO quantify_VVI the_AT popularity_NN1 of_IO a_AT1 particular_JJ detected_JJ site_NN1 ._. 
</s>
<s>
Agile_JJ EA_NP1 artefacts_NN2 are_VBR initially_RR modelled_VVN at_II the_AT high-level_JJ ._. 
</s>
<s>
For_REX21 example_REX22 ,_, when_CS released_VVN in_II 2011_MC ,_, concerns_VVZ over_II Snapchat_NP1 were_VBDR raised_VVN regarding_II the_AT '_NULL self-destruct_VV0 '_NULL nature_NN1 of_IO its_APPGE messages_NN2 ,_, and_CC it_PPH1 '_NULL s_ZZ1 potential_JJ for_IF distributing_VVG illicit_JJ material_NN1 (_( Poltash_NP1 ,_, 2013_MC )_) ._. 
</s>
<s>
Specifically_RR ,_, this_DD1 method_NN1 is_VBZ suitable_JJ for_IF multi-scans_NN2 shape_VV0 registration_NN1 with_IW little_DA1 overlap_NN1 ._. 
</s>
<s>
Each_DD1 Hessian-vector_JJ product_NN1 calculation_NN1 requires_VVZ one_MC1 adjoint_JJ model_NN1 run_NN1 and_CC one_MC1 tangent_JJ linear_JJ model_NN1 run_NN1 ._. 
</s>
<s>
Simulation_NN1 results_NN2 of_IO the_AT original_JJ micro-topography_NN1 model_NN1 ,_, presented_VVN in_II Frei_NP1 et_NP1 ?_? al_NP1 ._. 
</s>
<s>
Iverson_NP1 &lsqb;_( 8_MC &rsqb;_) and_CC later_RRR Iverson_NP1 and_CC Denlinger_NP1 &lsqb;_( 9_MC &rsqb;_) argue_VV0 that_CST interstitial_JJ fluid_NN1 alters_VVZ the_AT behavior_NN1 of_IO flows_NN2 and_CC must_VM be_VBI included_VVN in_II the_AT constitutive_JJ behavior_NN1 of_IO the_AT flowing_JJ material_NN1 ._. 
</s>
<s>
For_IF now_RT ,_, the_AT statistical_JJ error_NN1 would_VM definitely_RR be_VBI large_JJ ;_; with_IW a_AT1 larger_JJR sample_NN1 the_AT results_NN2 would_VM also_RR be_VBI improved_VVN in_II quantitative_JJ precision_NN1 ._. 
</s>
<s>
First_MD ,_, a_AT1 univariate_NN1 time_NNT1 series_NN analysis_NN1 to_TO identify_VVI what_DDQ variables_NN2 should_VM be_VBI included_VVN in_II the_AT multivariate_JJ model_NN1 ._. 
</s>
<s>
These_DD2 have_VH0 been_VBN chosen_VVN on_II the_AT basis_NN1 that_CST ,_, unlike_JJ various_JJ security-specific_JJ usability_NN1 guidelines_NN2 that_CST have_VH0 emerged_VVN from_II academic_JJ studies_NN2 (_( which_DDQ could_VM be_VBI argued_VVN to_TO be_VBI obscure_JJ and_CC immature_JJ at_II this_DD1 stage_NN1 )_) ,_, Nielsen_NP1 '_NULL s_ZZ1 recommendations_NN2 are_VBR well-known_JJ and_CC widely_RR cited_VVN ._. 
</s>
<s>
We_PPIS2 have_VH0 chosen_VVN not_XX to_TO implement_VVI the_AT interactions_NN2 between_II system_NN1 components_NN2 using_VVG Web_NN1 Services_NN2 technologies_NN2 such_II21 as_II22 SOAP_NN1 ._. 
</s>
<s>
Since_CS we_PPIS2 considered_VVD ideal_JJ queue_NN1 behavior_NN1 with_IW no_AT traffic_NN1 abnormality_NN1 ,_, the_AT behavior_NN1 pastWmaxis_NN1 not_XX visible_JJ ._. 
</s>
<s>
The_AT three_MC disjoint_JJ ,_, finite_JJ sets_NN2 of_IO such_DA a_AT1 graph_NN1 correspond_VV0 to_II (_( 1_MC1 )_) a_AT1 set_NN1 of_IO users_NN2 u∈U_FO ,_, (_( 2_MC )_) a_AT1 set_NN1 of_IO objects_NN2 or_CC resources_NN2 r∈R_FO and_CC (_( 3_MC )_) a_AT1 set_NN1 of_IO annotations_NN2 or_CC tags_VVZ t∈T_FO that_CST are_VBR used_VVN by_II users_NN2 U_JJ to_TO annotate_VVI resources_NN2 R._NP1 For_REX21 instance_REX22 ,_, for_IF a_AT1 problem_NN1 with_IW spatial_JJ variability_NN1 (_( e.g._REX ,_, sheet_NN1 metal_NN1 thickness_NN1 distribution_NN1 )_) ,_, one_PN1 should_VM choose_VVI to_TO describe_VVI the_AT problem_NN1 with_IW random_JJ fields_NN2 as_CSA they_PPHS2 provide_VV0 a_AT1 more_RGR realistic_JJ representation_NN1 than_CSN uncorrelated_JJ random_JJ variables_NN2 ._. 
</s>
<s>
DK_NP1 subsequently_RR responded_VVN communicating_VVG the_AT deficiency_NN1 of_IO AUTOnline_NP1 email_NN1 for_IF contacting_VVG Swedish_JJ students_NN2 ,_, and_CC indicating_VVG a_AT1 resolution_NN1 to_II the_AT problem_NN1 ,_, requiring_VVG an_AT1 active_JJ TUM_NN1 adjustment_NN1 activity_NN1 on_II both_DB2 the_AT student_NN1 and_CC DK_NP1 '_NULL s_ZZ1 part_NN1 ,_, since_CS DK_NP1 had_VHD individual_JJ email_NN1 addresses_VVZ for_IF Swedish_JJ students_NN2 ._. 
</s>
<s>
All_DB these_DD2 restricted_JJ forms_NN2 of_IO skill_NN1 games_NN2 have_VH0 concise_JJ representations_NN2 ,_, since_CS we_PPIS2 can_VM find_VVI a_AT1 succinct_JJ representation_NN1 for_IF the_AT task_NN1 value_NN1 function_NN1 ,_, and_CC thus_RR also_RR have_VH0 a_AT1 succinct_JJ representation_NN1 for_IF the_AT characteristic_JJ function_NN1 ._. 
</s>
<s>
Equivalent_JJ words_NN2 between_II the_AT two_MC target_NN1 word_NN1 lists_NN2 are_VBR put_VVN into_II the_AT returned_JJ collection_NN1 and_CC used_VVN as_II the_AT basis_NN1 for_IF service_NN1 recommendation_NN1 ._. 
</s>
<s>
Whether_CSW in_II the_AT control_NN1 or_CC experimental_JJ groups_NN2 ,_, subjects_NN2 had_VHD to_TO spend_VVI at_RR21 least_RR22 some_DD time_NNT1 on_II understanding_VVG the_AT program_NN1 requirements_NN2 ,_, on_II reading_VVG and_CC understanding_VVG the_AT program_NN1 comprehension_NN1 questions_NN2 ,_, and_CC on_II recording_VVG their_APPGE answers_NN2 ._. 
</s>
<s>
Fig._NN1 1_MC1 depicts_VVZ the_AT message-passing_JJ behaviour_NN1 of_IO a_AT1 customer_NN1 and_CC a_AT1 travel_NN1 agent_NN1 in_II a_AT1 travel_NN1 reservation_NN1 scenario_NN1 ._. 
</s>
<s>
The_AT first_MD aircraft_NN used_VVN in_II early_JJ air_NN1 warfare_NN1 did_VDD not_XX come_VVI with_IW a_AT1 set_NN1 of_IO air_NN1 warfare_NN1 principles_NN2 ,_, they_PPHS2 were_VBDR developed_VVN based_VVN on_II the_AT experiences_NN2 of_IO air_NN1 warfare_NN1 pioneers_NN2 ._. 
</s>
<s>
We_PPIS2 have_VH0 just_RR received_VVN disk_NN1 and_CC machines_NN2 to_TO handle_VVI roughly_RR that_DD1 amount_NN1 ._. 
</s>
<s>
Publication_NN1 venue_NN1 also_RR contributes_VVZ for_IF an_AT1 improvement_NN1 of_IO 0.01_MC on_II Scopus_NP1 and_CC 0.005_MC on_II DBLP_NP1 respectively_RR ._. 
</s>
<s>
It_PPH1 is_VBZ now_RT recognised_VVN that_CST during_II the_AT adoption_NN1 of_IO the_AT e-hub_JJ '_NULL limited_JJ buy-in_NN1 by_II stakeholders_NN2 &lsqb;_( during_II &rsqb;_) registration_NN1 has_VHZ been_VBN a_AT1 nightmare_NN1 ._. 
</s>
<s>
It_PPH1 was_VBDZ recognized_VVN that_CST it_PPH1 is_VBZ not_XX the_AT purpose_NN1 of_IO a_AT1 provenance_NN1 standardization_NN1 activity_NN1 to_TO specify_VVI them_PPHO2 ._. 
</s>
<s>
Is_VBZ it_PPH1 ,_, for_REX21 instance_REX22 ,_, possible_JJ that_CST a_AT1 particular_JJ person_NN1 has_VHZ several_DA2 stretches_NN2 of_IO behaviour_NN1 going_VVG on_RP in_II parallel_NN1 during_II some_DD given_JJ period_NN1 of_IO time_NNT1 ?_? 
</s>
<s>
It_PPH1 can_VM be_VBI accessed_VVN immediately_RR through_II any_DD suitable_JJ browser_NN1 ,_, and_CC new_JJ versions_NN2 become_VV0 accessible_JJ as_CS31 soon_CS32 as_CS33 they_PPHS2 are_VBR released_VVN ._. 
</s>
<s>
We_PPIS2 therefore_RR do_VD0 not_XX include_VVI the_AT IntersectionSubTree_NN1 kernels_NN2 ,_, which_DDQ do_VD0 not_XX have_VHI such_DA a_AT1 representation_NN1 ._. 
</s>
<s>
The_AT algorithm_NN1 estimates_VVZ this_DD1 quantity_NN1 by_II finding_VVG the_AT most_RGT likely_JJ compound_NN1 event_NN1 ,_, made_VVD up_RP of_IO a_AT1 certain_JJ number_NN1 of_IO long_JJ bursts_NN2 ,_, of_IO a_AT1 certain_JJ length_NN1 ,_, in_II combination_NN1 with_IW a_AT1 certain_JJ quantity_NN1 of_IO traffic_NN1 from_II the_AT short_JJ bursts_NN2 ,_, to_TO cause_VVI an_AT1 overflow_NN1 ._. 
</s>
<s>
If_CS and_CC when_CS the_AT execution_NN1 of_IO the_AT solution_NN1 for_IF D_ZZ1 fails_VVZ in_II the_AT probabilistic_JJ environment_NN1 ,_, FF_NNU is_VBZ reinvoked_VVN to_TO plan_VVI again_RT from_II the_AT failed_JJ state_NN1 ._. 
</s>
<s>
In_II the_AT illustrative_JJ example_NN1 ,_, WM_NP1 =_FO Rec_NP1 ,_, NRec_NP1 and_CC the_AT composition_NN1 of_IO the_AT generated_JJ mass_NN1 found_VVN in_II Table_NN1 2_MC is_VBZ used_VVN to_TO calculate_VVI β_NULL wm_NNU ,_, t_ZZ1 for_IF each_DD1 material_NN1 in_II each_DD1 stage_NN1 ._. 
</s>
<s>
In_II this_DD1 study_NN1 and_CC in_II my_APPGE discussion_NN1 I_PPIS1 have_VH0 often_RR separated_VVN the_AT social_JJ and_CC material_NN1 issues_NN2 analytically_RR ;_; in_II practice_NN1 for_IF the_AT informants_NN2 they_PPHS2 were_VBDR not_XX distinct_JJ issues_NN2 ._. 
</s>
<s>
Thus_RR this_DD1 reengineering_NN1 has_VHZ no_AT impact_NN1 on_II the_AT network_NN1 side_NN1 dynamics_NN of_IO the_AT protocol_NN1 ._. 
</s>
<s>
Location-based_JJ information_NN1 has_VHZ rarely_RR been_VBN used_VVN in_II transactional_JJ public_JJ services_NN2 in_II Flanders_NP1 because_CS a_AT1 lack_NN1 of_IO vocabularies_NN2 that_CST bridge_NN1 between_II alphanumeric_JJ and_CC geographical_JJ information_NN1 ._. 
</s>
<s>
Given_CS21 that_CS22 there_EX are_VBR daily_JJ and_CC weekly_JJ backup_NN1 processes_NN2 ,_, a_AT1 period_NN1 of_IO one_MC1 day_NNT1 would_VM be_VBI too_RG short_JJ ._. 
</s>
<s>
The_AT utility_NN1 and_CC limitations_NN2 of_IO DUET-H/WQ_NP1 were_VBDR explored_VVN by_II its_APPGE application_NN1 in_II several_DA2 case_NN1 studies_NN2 with_IW real-world_JJ monitoring_NN1 data_NN ._. 
</s>
<s>
LetPfbe_VV0 the_AT set_NN1 of_IO all_DB routing_VVG matrices∈_FO &lsqb;_( 0,1_MC &rsqb;_) (_( S-1_FO )_) 2_MC ,_, for_IF which_DDQ the_AT corresponding_JJ total_JJ link_NN1 flows_VVZ satisfyfij<Cijfor_FO all_RR (_( i_ZZ1 ,_, j_ZZ1 )_) ∈A_FO ._. 
</s>
<s>
Ipsative_JJ metrics_NN2 are_VBR often_RR used_JJ to_II measure_NN1 subjective_JJ interpretation_NN1 of_IO psychological_JJ constructs_NN2 (_( van_NP1 Eijnatten_NP1 et_RA21 al._RA22 ,_, 2014_MC )_) ._. 
</s>
<s>
During_II early_JJ system_NN1 design_NN1 the_AT system_NN1 being_VBG modeled_VVN is_VBZ not_XX com_VV0 letely_RR known_VVN ,_, Moreover_RR ,_, due_II21 to_II22 the_AT integration_NN1 of_IO optimised_JJ UDF_NP1 s_ZZ1 from_II ExaStream_NP1 (_( such_II21 as_II22 an_AT1 optimised_JJ version_NN1 of_IO the_AT correlation_NN1 function_NN1 )_) ,_, STARQL_NP1 offers_VVZ the_AT main_JJ components_NN2 for_IF an_AT1 analytics_NN1 aware_JJ OBDAapproach_NN1 as_CSA described_VVN in_II &lsqb;_( 43_MC &rsqb;_) ._. 
</s>
<s>
Note_VV0 that_CST the_AT goal_NN1 of_IO this_DD1 work_NN1 is_VBZ not_XX to_TO present_VVI another_DD1 model_NN1 selection_NN1 method_NN1 ,_, nor_CC to_TO compare_VVI existing_JJ criterion-based_JJ selection_NN1 methods_NN2 to_II one_PPX121 another_PPX122 (_( as_CSA has_VHZ been_VBN previously_RR done_VDN in_II ,_, e.g._REX Cherkassky_NP1 and_CC Ma_NP1 ,_, 2003_MC ;_; Hastie_NP1 et_NP1 ?_? al._NNU ,_, 2001_MC )_) but_CCB rather_RR to_TO dramatically_RR decrease_VVI the_AT computational_JJ cost_NN1 of_IO implementing_VVG existing_JJ criterion-based_JJ methods_NN2 ,_, especially_RR for_IF cases_NN2 where_RRQ the_AT number_NN1 (_( r_ZZ1 )_) of_IO candidate_NN1 covariates_NN2 is_VBZ large_JJ and/or_CC where_RRQ regression_NN1 residuals_NN2 are_VBR suspected_VVN of_IO being_VBG spatially_RR and/or_CC temporally_RR auto-correlated_VVN ._. 
</s>
<s>
An_AT1 advantage_NN1 of_IO the_AT frequency-domain_JJ methods_NN2 is_VBZ that_CST ,_, in_RR21 general_RR22 ,_, they_PPHS2 are_VBR robust_JJ to_II malicious_JJ attacks_NN2 ._. 
</s>
<s>
This_DD1 provides_VVZ support_NN1 for_IF a_AT1 positive_JJ answer_NN1 to_II RQ3_FO ,_, showing_VVG that_CST an_AT1 addition_NN1 of_IO manually_RR annotated_VVN real_JJ world_NN1 examples_NN2 can_VM actually_RR drive_VVI the_AT model_NN1 to_II a_AT1 better_JJR comprehension_NN1 of_IO the_AT actual_JJ problem_NN1 space_NN1 sampled_VVD ,_, in_II our_APPGE case_NN1 ,_, by_II the_AT 425M_NNU reference_NN1 set_NN1 ._. 
</s>
<s>
The_AT proposed_JJ utility_NN1 function_NN1 calculates_VVZ the_AT rank_NN1 of_IO features_NN2 based_VVN on_II their_APPGE impact_NN1 on_II non-functional_JJ properties_NN2 by_II considering_VVG the_AT preferences_NN2 of_IO stakeholders_NN2 formulated_VVN in_II31 terms_II32 of_II33 the_AT weight_NN1 of_IO non-functional_JJ properties_NN2 ._. 
</s>
<s>
We_PPIS2 recall_VV0 that_CST (_( in_II two_MC dimensions_NN2 )_) the_AT velocity_NN1 components_NN2 are_VBR simply_RR first_MD derivatives_NN2 of_IO the_AT stream_NN1 function_NN1 (_( components_NN2 of_IO the_AT curl_NN1 )_) ._. 
</s>
<s>
We_PPIS2 find_VV0 that_CST higher_JJR levels_NN2 of_IO Internet_NP1 purchasing_NN1 are_VBR associated_VVN with_IW higher_JJR levels_NN2 of_IO total_JJ outsourcing_NN1 (_( p@<@.05_FO )_) ._. 
</s>
<s>
Focal_JJ habitat_NN1 edge_NN1 interfaces_NN2 are_VBR differentiated_VVN by_II morphology_NN1 (_( edges_NN2 of_IO interior_NN1 ,_, connector_NN1 and_CC branch_NN1 as_CSA linear_JJ features_NN2 ,_, islets_NN2 )_) and_CC characterised_VVN according_II21 to_II22 the_AT similarity_NN1 of_IO adjacent_JJ habitats_NN2 (_( focal_JJ habitat_NN1 edges_NN2 along_II natural/semi-natural_JJ lands_NN2 as_CSA natural_JJ edge_NN1 interface_NN1 ,_, edges_NN2 along_II more_RGR anthropogenic_JJ (_( i.e._REX agricultural_JJ and/or_CC artificial_JJ )_) lands_NN2 as_CSA artificial_JJ edge_NN1 interface_NN1 )_) ._. 
</s>
<s>
The_AT final_JJ time_NNT1 varies_VVZ from_II 50_MC to_II 600_MC and_CC we_PPIS2 calculate_VV0 the_AT solution_NN1 at_II tf_NNU =_FO 50_MC ,_, 100_MC ,_, 200_MC ,_, ,_, 600_MC ._. 
</s>
<s>
Such_DA information_NN1 discovery_NN1 and_CC visualisation_NN1 functionalities_NN2 are_VBR provided_VVN by_II GATE_NN1 Prospector_NN1 &lsqb;_( 13_MC &rsqb;_) ,_, which_DDQ is_VBZ a_AT1 web-based_JJ user_NN1 interface_NN1 for_IF searching_VVG and_CC visualising_VVG correlations_NN2 in_II large_JJ datasets_NN2 ._. 
</s>
<s>
However_RR ,_, this_DD1 is_VBZ not_XX the_AT case_NN1 for_IF all_DB software_NN1 engineering_NN1 mapping_NN1 studies_NN2 ._. 
</s>
<s>
It_PPH1 identifies_VVZ the_AT parameters_NN2 in_II the_AT semantic_JJ service_NN1 description_NN1 associated_VVN to_II the_AT process_NN1 (_( line_NN1 49_MC and_CC 50_MC )_) ._. 
</s>
<s>
If_CS a_AT1 router_NN1 finds_VVZ its_APPGE RPF_NP1 check_NN1 succeeds_VVZ ,_, it_PPH1 keeps_VVZ a_AT1 copy_NN1 of_IO the_AT packet_NN1 and_CC forwards_RL the_AT packet_NN1 and_CC the_AT respective_JJ reduced_VVD Pd_NP1 values_NN2 further_RRR to_II all_DB its_APPGE neighbors_NN2 within_II its_APPGE received_JJ Pd_NP1 range_NN1 except_CS to_II the_AT sender_NN1 of_IO the_AT packet_NN1 ._. 
</s>
<s>
Combined_VVN with_IW security_NN1 patterns_NN2 (_( &lsqb;_( Cheng_NP1 et@al._FO ,_, 2003_MC &rsqb;_) and_CC &lsqb;_( Weiss_NP1 and_CC Mouratidis_NP1 ,_, 2008_MC &rsqb;_) )_) ,_, these_DD2 techniques_NN2 focus_VV0 elicitation_NN1 on_II identifying_VVG security_NN1 vulnerabilities_NN2 from_II which_DDQ mitigating_JJ security_NN1 requirements_NN2 are_VBR derived_VVN ._. 
</s>
<s>
However_RR ,_, in_II these_DD2 protocols_NN2 the_AT end-hosts_NN2 are_VBR considered_VVN to_TO be_VBI equivalent_JJ peers_NN2 and_CC are_VBR organized_VVN into_II an_AT1 appropriate_JJ overlay_NN1 structure_NN1 for_IF multicast_NN1 data_NN delivery_NN1 ._. 
</s>
<s>
Hence_RR ,_, it_PPH1 would_VM be_VBI better_JJR to_TO control_VVI the_AT phosphorous_JJ at_II the_AT watershed_NN1 level_NN1 than_CSN to_TO build_VVI a_AT1 plant_NN1 ._. 
</s>
<s>
Given_VVN a_AT1 network_NN1 with_IW certain_JJ properties_NN2 ,_, what_DDQ are_VBR the_AT best_JJT ways_NN2 to_TO search_VVI for_IF particular_JJ nodes_NN2 in_II this_DD1 network_NN1 ?_? 
</s>
<s>
An_AT1 aggregated_JJ cloud_NN1 service_NN1 adoption_NN1 rate_NN1 (_( AR_UH )_) was_VBDZ calculated_VVN as_II a_AT1 cloud_NN1 infrastructure_NN1 would_VM deliver_VVI all_DB services_NN2 through_II a_AT1 single_JJ infrastructure_NN1 ._. 
</s>
<s>
This_DD1 problem_NN1 has_VHZ been_VBN investigated_VVN by_II other_JJ researchers_NN2 &lsqb;_( 39_MC &rsqb;_) ,_, &lsqb;_( 14_MC &rsqb;_) ,_, &lsqb;_( 15_MC &rsqb;_) and&lsqb;16&rsqb;_FO using_VVG both_RR Finite_JJ Elements_NN2 and_CC two_MC alternative_JJ formulations_NN2 of_IO Particle_NN1 Finite_JJ Elements_NN2 methods_NN2 ._. 
</s>
<s>
Shape_NN1 representation_NN1 using_VVG Fourier_NP1 descriptors_NN2 is_VBZ easy_JJ to_TO compute_VVI and_CC robust_JJ ._. 
</s>
<s>
The_AT derived_JJ discriminant_JJ analysis_NN1 method_NN1 ,_, LM-NNDA_NP1 ,_, is_VBZ the_AT most_RGT suitable_JJ feature_NN1 extractor_NN1 for_IF the_AT LM-NN_JJ classifier_NN1 in_II theory_NN1 ._. 
</s>
<s>
The_AT coordinated_JJ attacks_NN2 described_VVN above_RL are_VBR summarized_VVN inTable_JJ 1_MC1 ._. 
</s>
<s>
There_EX are_VBR a_AT1 variety_NN1 of_IO approaches_NN2 to_II applying_VVG SPE_NN1 practices_NN2 to_II distributed_JJ real-time_JJ systems_NN2 (_( e.g._REX ,_, &lsqb;_( 17_MC &rsqb;_) and&lsqb;18&rsqb;_FO )_) ._. 
</s>
<s>
Albrechtsen_NP1 (_( 2007_MC )_) conducted_VVD an_AT1 interview_NN1 study_NN1 to_TO explain_VVI users_NN2 '_NULL experiences_NN2 of_IO information_NN1 security_NN1 by_II organizational_JJ factors_NN2 ._. 
</s>
<s>
Five_MC of_IO those_DD2 studies_NN2 found_VVN in_II31 favour_II32 of_II33 expert-based_JJ methods_NN2 ,_, five_MC found_VVD no_AT difference_NN1 ,_, and_CC five_MC found_VVD in_II31 favour_II32 of_II33 model-based_JJ estimation_NN1 ._. 
</s>
<s>
The_AT relationship_NN1 shows_VVZ that_CST the_AT motion_NN1 field_NN1 depends_VVZ on_II the_AT distribution_NN1 of_IO depths_NN2 F_ZZ1 (_( x_ZZ1 ,_, y_ZZ1 )_) ._. 
</s>
<s>
Most_DAT subjects_NN2 (_( 77%_NNU )_) indicated_VVD English_NN1 was_VBDZ their_APPGE primary_JJ language_NN1 ._. 
</s>
<s>
Then_RT we_PPIS2 demonstrate_VV0 the_AT applicability_NN1 of_IO the_AT semantic_JJ approach_NN1 by_II31 means_II32 of_II33 the_AT implementation_NN1 of_IO two_MC semantic_JJ services_NN2 :_: the_AT data_NN acquisition_NN1 service_NN1 and_CC the_AT data_NN analysis_NN1 service_NN1 ._. 
</s>
<s>
Trees_NN2 are_VBR simulated_VVN as_CSA adaptive_JJ agents_NN2 that_CST compete_VV0 for_IF resources_NN2 (_( predominately_RR light_JJ ,_, but_CCB also_RR water_NN1 and_CC nutrients_NN2 are_VBR considered_VVN )_) ,_, and_CC dynamically_RR adapt_VV0 to_II their_APPGE environment_NN1 (_( Seidl_NP1 et_NP1 ?_? al._NNU ,_, 2012a_FO )_) ._. 
</s>
<s>
Tokens_NN2 placed_VVN at_II wf_NNU are_VBR evaluated_VVN in_II parallel_NN1 by_II the_AT BelongToTuple_NN1 transitions_NN2 ._. 
</s>
<s>
While_CS Oncosimulator_NP1 itself_PPX1 is_VBZ not_XX a_AT1 unified_JJ environment_NN1 for_IF data_NN sharing_VVG and_CC analysis_NN1 ,_, it_PPH1 will_VM be_VBI integrated_VVN with_IW the_AT IMENSE_NN1 system_NN1 via_II a_AT1 workflow_NN1 ._. 
</s>
<s>
However_RR ,_, if_CS there_EX are_VBR at_RR21 most_RR22 three_MC candidates_NN2 ,_, the_AT computational_JJ complexity_NN1 of_IO computing_VVG a_AT1 manipulation_NN1 under_II Nanson_NP1 '_NULL s_ZZ1 rule_NN1 is_VBZ polynomial-time_JJ ._. 
</s>
<s>
One_MC1 of_IO the_AT potential_JJ advantages_NN2 of_IO some_DD biometric_JJ options_NN2 is_VBZ that_CST they_PPHS2 can_VM be_VBI utilised_VVN non-intrusively_RR and_CC continuously_RR (_( or_CC at_II discrete_JJ intervals_NN2 )_) during_II the_AT user_NN1 session_NNT1 ._. 
</s>
<s>
Switching_JJ costs_NN2 are_VBR associated_VVN with_IW the_AT likelihood_NN1 of_IO continuing_VVG an_AT1 exchange_NN1 relationship_NN1 with_IW a_AT1 supplier_NN1 (_( Weiss_NP1 and_CC Anderson_NP1 ,_, 1992_MC ,_, Ping_NN1 ,_, 1993_MC and_CC Morgan_NP1 and_CC Hunt_NP1 ,_, 1994_MC )_) and_CC customer_NN1 repurchase_NN1 intentions_NN2 (_( Jones_NP1 et_RA21 al._RA22 ,_, 2002_MC )_) ._. 
</s>
<s>
Using_VVG the_AT above_JJ sets_NN2 of_IO benchmarks_NN2 in_II two_MC categories_NN2 :_: simple_JJ and_CC optimised_JJ benchmarks_NN2 ,_, we_PPIS2 can_VM estimate_VVI concrete_JJ times_NNT2 for_IF our_APPGE protocol_NN1 on_II the_AT device_NN1 side_NN1 ._. 
</s>
<s>
Our_APPGE program_NN1 will_VM help_VVI fill_VVI those_DD2 needs_NN2 in_II Canada_NP1 ._. 
</s>
<s>
With_IW this_DD1 in_II mind_NN1 ,_, we_PPIS2 can_VM assume_VVI that_CST the_AT robustness_NN1 of_IO this_DD1 method_NN1 minimizes_VVZ the_AT risks_NN2 of_IO drawing_VVG conclusions_NN2 from_II poor_JJ data_NN and_CC perhaps_RR highlights_VVZ what_DDQ information_NN1 should_VM be_VBI considered_VVN most_DAT relevant_JJ to_TO ultimately_RR refining_VVG the_AT predictive_JJ power_NN1 ._. 
</s>
<s>
This_DD1 select_JJ process_NN1 was_VBDZ purposely_RR constructed_VVN to_TO return_VVI an_AT1 in-tree_JJ path_NN1 that_CST is_VBZ preferred_VVN by_II its_APPGE initial_JJ edge_NN1 weight_NN1 (_( to_II some_DD specific_JJ next-hop_JJ neighbor_NN1 )_) rather_CS21 than_CS22 by_II the_AT path_NN1 '_NULL s_ZZ1 length_NN1 ._. 
</s>
<s>
Since_CS the_AT deformation_NN1 kinematics_NN2 are_VBR specified_VVN ,_, it_PPH1 excludes_VVZ the_AT need_NN1 to_TO solve_VVI the_AT elastic_JJ response_NN1 of_IO the_AT body_NN1 and_CC the_AT fins_NN2 ._. 
</s>
<s>
The_AT decision_NN1 stage_NN1 is_VBZ coupled_VVN to_II the_AT modeling/monitoring_NN1 stage_NN1 by_II the_AT system_NN1 conceptualization_NN1 ,_, which_DDQ represents_VVZ a_AT1 high_JJ level_JJ view_NN1 of_IO the_AT system_NN1 within_II which_DDQ the_AT problem_NN1 occurs_VVZ ._. 
</s>
<s>
Thus_RR ,_, the_AT resulting_JJ task_NN1 set_NN1 is_VBZ uniform_JJ ._. 
</s>
<s>
In_II this_DD1 sense_NN1 we_PPIS2 see_VV0 our_APPGE approach_NN1 as_II a_AT1 complement_NN1 to_II current_JJ state-of-the-art_JJ matching_JJ techniques_NN2 as_CSA it_PPH1 may_VM provide_VVI valuable_JJ information_NN1 for_IF pruning_VVG the_AT search_NN1 space_NN1 or_CC disambiguating_VVG the_AT results_NN2 of_IO candidate_NN1 semantic_JJ alignments_NN2 computed_VVN with_IW today_RT '_NULL s_ZZ1 ontology-matching_JJ technology_NN1 ._. 
</s>
<s>
Thus_RR ,_, any_DD one_MC1 form_NN1 with_IW scalar_JJ coefficients_NN2 has_VHZ the_AT unique_JJ component_NN1 representation_NN1 with_II31 respect_II32 to_II33 the_AT standard_JJ basis_NN1 ,_, =idxi+tdt=idxi+tdt_FO ,_, in_II which_DDQ ii_MC and_CC tt_UH are_VBR scalar_JJ fields_NN2 on_II DD_NNA ._. 
</s>
<s>
One_MC1 participant_NN1 stated_VVD that_CST the_AT two-factor_JJ version_NN1 "_" was_VBDZ too_RG much_DA1 effort_NN1 ._. 
</s>
<s>
The_AT second_MD type_NN1 of_IO email_NN1 was_VBDZ the_AT HTML_NP1 attachment_NN1 to_II an_AT1 email_NN1 ._. 
</s>
<s>
The_AT report_NN1 example_NN1 portrays_VVZ the_AT top_NN1 25_MC rated_JJ configurations_NN2 ._. 
</s>
<s>
The_AT plots_NN2 show_VV0 the_AT number_NN1 of_IO states_NN2 evaluated_VVN in_II the_AT four_MC different_JJ configurations_NN2 of_IO the_AT heuristic_JJ for_IF the_AT different_JJ load_NN1 profiles_NN2 and_CC for_IF problems_NN2 with_IW an_AT1 increasing_JJ number_NN1 of_IO capacitors_NN2 ._. 
</s>
<s>
It_PPH1 was_VBDZ noted_VVN that_CST reviewing_VVG the_AT test_NN1 cases_NN2 helps_VVZ development_NN1 stay_VVI on_II track_NN1 in_II an_AT1 iteration_NN1 because_CS "_" if_CS the_AT tests_NN2 are_VBR created_VVN based_VVN on_II specification?it_NN1 '_NULL s_ZZ1 a_AT1 safety_NN1 net_NN1 that_CST QA_NN1 can_VM actually_RR find_VVI out_RP if_CS it_PPH1 '_NULL s_ZZ1 correct_JJ or_CC not_XX "_" (_( BA_NP1 ,_, Case_NN1 3_MC )_) ._. 
</s>
<s>
GPU_NP1 processing_NN1 ,_, especially_RR using_VVG lower_JJR level_NN1 kernels_NN2 ,_, can_VM offer_VVI another_DD1 quantum_NN1 leap_NN1 in_II performance_NN1 ._. 
</s>
<s>
This_DD1 section_NN1 reviews_VVZ the_AT BGP_NP1 protocol_NN1 and_CC discusses_VVZ two_MC important_JJ classes_NN2 of_IO BGP_NP1 attacks_NN2 ,_, origin_NN1 AS_CSA attacks_NN2 and_CC invalid_JJ paths_NN2 ._. 
</s>
<s>
The_AT increasing_JJ requirement_NN1 for_IF protection_NN1 is_VBZ evidenced_VVN by_II a_AT1 survey_NN1 of_IO 230_MC business_NN1 professionals_NN2 ,_, which_DDQ found_VVD that_CST 81%_NNU considered_VVD the_AT information_NN1 on_II their_APPGE PDA_NN1 was_VBDZ either_RR somewhat_RR or_CC extremely_RR valuable_JJ ._. 
</s>
<s>
Alignment_NN1 between_II the_AT ETS_NP2 data_NN structure_NN1 and_CC scenarios_NN2 data_NN structures_NN2 ._. 
</s>
<s>
Automatic_JJ generation_NN1 of_IO code_NN1 from_II different_JJ kinds_NN2 of_IO models_NN2 has_VHZ increased_VVN in_II popularity_NN1 as_II the_AT technology_NN1 has_VHZ become_VVN more_RGR robust_JJ ._. 
</s>
<s>
Here_RL too_RR we_PPIS2 can_VM require_VVI either_RR a_AT1 small_JJ relative_JJ error_NN1 or_CC a_AT1 small_JJ absolute_JJ error_NN1 ._. 
</s>
<s>
Adopting_VVG a_AT1 new_JJ algorithmic_JJ approach_NN1 to_II runtime_NNT1 plan_NN1 analysis_NN1 to_TO better_RRR take_VVI advantage_NN1 of_IO opportunities_NN2 for_IF parallelization_NN1 ._. 
</s>
<s>
There_EX are_VBR two_MC sets_NN2 of_IO variables_NN2 in_II the_AT calculation_NN1 of_IO this_DD1 cost_NN1 ._. 
</s>
<s>
This_DD1 first_MD image_NN1 is_VBZ marked_VVN as_II the_AT reference_NN1 image_NN1 of_IO the_AT plane_NN1 ,_, Irm_NP1 ._. 
</s>
<s>
Also_RR ,_, new_JJ Northbound_NP1 APIs_NP1 need_VV0 to_TO be_VBI specified_VVN that_CST will_VM expose_VVI a_AT1 uniform_JJ configuration_NN1 interface_NN1 through_II which_DDQ the_AT OpenADN_NN1 data_NN plane_NN1 entities_NN2 may_VM be_VBI configured_VVN ._. 
</s>
<s>
The_AT main_JJ difference_NN1 between_II stSPARQL_NN1 and_CC GeoSPARQL_NN1 is_VBZ that_CST stSPARQL_VV0 also_RR provides_VVZ support_NN1 spatial_JJ updates_NN2 and_CC spatial_JJ aggregates_NN2 ,_, and_CC offer_VV0 valid_JJ time_NNT1 support_NN1 ._. 
</s>
<s>
Moreover_RR ,_, somewhat_RR surprisingly_RR ,_, it_PPH1 turns_VVZ out_RP that_DD1 reasoning_NN1 about_II minimal_JJ knowledge_NN1 is_VBZ harder_JJR than_CSN reasoning_VVG about_RP only_RR knowing_VVG :_: see_VV0 &lsqb;_( 21_MC &rsqb;_) for_IF a_AT1 complexity_NN1 analysis_NN1 (_( in_II the_AT propositional_JJ setting_NN1 wrt_NNU a_AT1 single_JJ agent_NN1 )_) ._. 
</s>
<s>
This_DD1 section_NN1 is_VBZ dedicated_VVN to_II the_AT overview_NN1 of_IO spatially_RR explicit_JJ ecosystem_NN1 simulation_NN1 framework_NN1 as_II31 well_II32 as_II33 related_JJ software_NN1 design_NN1 issues_NN2 related_VVN to_II each_DD1 major_JJ simulation_NN1 step_NN1 ._. 
</s>
<s>
Nevertheless_RR ,_, the_AT multi-scale_JJ line_NN1 detectors_NN2 outlined_VVN and_CC evaluated_VVN in_II this_DD1 paper_NN1 are_VBR able_JK to_TO detect_VVI tracks_NN2 with_IW varying_JJ widths_NN2 and_CC are_VBR therefore_RR applicable_JJ to_II the_AT detection_NN1 of_IO tracks_NN2 with_IW a_AT1 single-pixel_JJ width_NN1 and_CC those_DD2 with_IW greater_JJR widths_NN2 ._. 
</s>
<s>
SuperMUC_NN1 is_VBZ an_AT1 IBM_NP1 System_NN1 x_ZZ1 iDataPlex_VV0 machine_NN1 with_IW 147,456_MC compute_VV0 cores_NN2 and_CC a_AT1 total_JJ peak_NN1 performance_NN1 of_IO 3.185_MC PFLOP/s_FU (_( 21.6_MC GFLOP/s_FU per_II core_NN1 )_) ._. 
</s>
<s>
First_MD ,_, all_DB the_AT concepts_NN2 are_VBR tagged_VVN with_IW their_APPGE semantic_JJ types_NN2 ,_, then_RT ,_, such_DA semantic_JJ types_NN2 together_RL with_IW a_AT1 minimal_JJ set_NN1 of_IO linguistic_JJ features_NN2 ,_, are_VBR used_VVN to_TO train_VVI and_CC evaluate_VVI different_JJ classifiers_NN2 ._. 
</s>
<s>
The_AT transport_NN1 protocol_NN1 is_VBZ assumed_VVN to_TO be_VBI a_AT1 connectionless_JJ transport_NN1 protocol_NN1 (_( i.e._REX ,_, UDP_NP1 )_) ._. 
</s>
<s>
For_REX21 example_REX22 ,_, searching_VVG for_IF "_" text_NN1 "_" returns_VVZ four_MC pages_NN2 of_IO results_NN2 (_( about_RG eighty_MC hits_NN2 )_) and_CC searching_VVG for_IF "_" phone_NN1 "_" returns_VVZ about_RG 40_MC hits_NN2 ._. 
</s>
<s>
In_II &lsqb;_( 33_MC &rsqb;_) ,_, &lsqb;_( 4_MC &rsqb;_) ,_, Alon_NP1 et_RA21 al_RA22 ._. 
</s>
<s>
proposed_VVN a_AT1 unified_JJ framework_NN1 for_IF spatiotemporal_JJ gesture_NN1 segmentation_NN1 and_CC recognition_NN1 ._. 
</s>
<s>
In_II the_AT planning_NN1 approach_NN1 the_AT same_DA plan_NN1 is_VBZ applied_VVN to_II the_AT set_NN1 of_IO 500_MC instances_NN2 for_IF each_DD1 of_IO the_AT 12_MC load_NN1 profiles_NN2 ._. 
</s>
<s>
Lamandé_NP1 (_( 2010_MC )_) reports_VVZ that_CST the_AT GrIDSure_NN1 authentication_NN1 system_NN1 (_( http_NNU :_: //www.gridsure.com_FU )_) has_VHZ been_VBN integrated_VVN into_II Microsoft_NP1 '_NULL s_ZZ1 Unified_JJ Access_NN1 Gateway_NP1 (_( UAG_NP1 )_) platform_NN1 ._. 
</s>
<s>
We_PPIS2 utilised_VVD anhp-finite_JJ element_NN1 discretization_NN1 and_CC presented_VVD an_AT1 improved_JJ block_NN1 Jacobi_NN1 preconditioner_NN1 ,_, which_DDQ treats_VVZ gradient_NN1 blocks_NN2 in_II a_AT1 more_RGR natural_JJ way_NN1 ,_, as_CSA a_AT1 novel_JJ contribution_NN1 ._. 
</s>
<s>
Our_APPGE study_NN1 shows_VVZ that_CST the_AT efficiency_NN1 of_IO performance_NN1 of_IO a_AT1 type_NN1 of_IO routing_NN1 ,_, measured_VVN as_II an_AT1 average_JJ value_NN1 of_IO the_AT NPT_NP1 time_NNT1 series_NN and_CC the_AT stability_NN1 measured_VVN as_II a_AT1 variability_NN1 of_IO these_DD2 series_NN ,_, changes_NN2 with_IW source_NN1 load_NN1 values_NN2 among_II the_AT considered_JJ ecfs_NN2 ._. 
</s>
<s>
The_AT results_NN2 of_IO the_AT evaluation_NN1 indicate_VV0 that_CST the_AT dispersion_NN1 algorithm_NN1 performs_VVZ adequately_RR to_TO estimate_VVI downwind_JJ concentrations_NN2 with_IW 84%_NNU of_IO the_AT estimates_NN2 within_II a_AT1 factor_NN1 of_IO two_MC of_IO the_AT observations_NN2 ,_, and_CC an_AT1 overall_JJ bias_NN1 of_IO 2%_NNU ._. 
</s>
<s>
When_CS the_AT task_NN1 is_VBZ no_RR21 longer_RR22 blocked_VVN ,_, it_PPH1 moves_VVZ to_II the_AT ready_JJ state_NN1 ,_, or_CC it_PPH1 moves_VVZ directly_RR to_II the_AT running_JJ state_NN1 by_II preempting_VVG the_AT currently_RR running_JJ task_NN1 if_CS it_PPH1 is_VBZ the_AT highest_JJT priority_NN1 task_NN1 ._. 
</s>
<s>
The_AT structured_JJ name_NN1 is_VBZ stored_VVN in_II the_AT form_NN1 of_IO xml_NNU elements_NN2 within_II the_AT TEI_NP1 header_NN1 title_NN1 element_NN1 (_( we_PPIS2 modified_VVD the_AT schema_NN1 to_TO accommodate_VVI this_DD1 additional_JJ metadata_NN1 )_) ._. 
</s>
<s>
It_PPH1 turns_VVZ out_RP that_CST this_DD1 expectation_NN1 is_VBZ wrong_JJ on_II both_DB2 accounts_NN2 ._. 
</s>
<s>
For_IF a_AT1 more_RGR specific_JJ assessment_NN1 ,_, Table_NN1 1_MC1 presents_VVZ an_AT1 evaluation_NN1 of_IO the_AT security_NN1 features_NN2 of_IO the_AT two_MC applications_NN2 against_II Nielsen_NP1 '_NULL s_ZZ1 usability_NN1 heuristics_NN2 (_( Nielsen_NP1 ,_, 2005_MC )_) ._. 
</s>
<s>
When_CS SS-AS_NP2 receives_VVZ the_AT INVITE_VV0 request_VV0 it_PPH1 initiates_VVZ a_AT1 new_JJ session_NNT1 and_CC contacts_NN2 the_AT Database_NN1 Manager_NN1 to_TO update_VVI the_AT user_NN1 '_NULL s_ZZ1 status_NN1 and_CC store_NN1 session_NNT1 information_NN1 such_II21 as_II22 the_AT MH_NP1 '_NULL s_ZZ1 current_JJ IP_NN1 address_NN1 and_CC port_NN1 ._. 
</s>
<s>
In_II31 line_II32 with_II33 instantiating_VVG ADAPTATION_NN1 and_CC STATE_NN1 ,_, which_DDQ require_VV0 event_NN1 trigger_NN1 points_VVZ to_TO affect_VVI changes_NN2 ,_, "_" any_DD update_NN1 of_IO subject_NN1 or_CC object_VV0 attributes_NN2 and_CC any_DD change_NN1 of_IO system_NN1 (_( environment_NN1 )_) conditions_NN2 triggers_VVZ the_AT re-evaluation_NN1 of_IO the_AT policy_NN1 by_II the_AT PDP_NP1 according_II21 to_II22 the_AT ongoing_JJ usage_NN1 session_NNT1 "_" (_( see_VV0 the_AT SECURITY_NN1 SESSION_NNT1 modifier_NN1 )_) "_" and_CC may_VM result_VVI in_II revocation_NN1 of_IO the_AT ongoing_JJ usage_NN1 or_CC update_NN1 of_IO attributes_NN2 if_CS necessary_JJ "_" (_( Zhang_NP1 et_RA21 al._RA22 ,_, 2008_MC )_) ._. 
</s>
<s>
This_DD1 voting_NN1 mechanism_NN1 accumulates_VVZ evidences_NN2 in_II the_AT locations_NN2 where_CS we_PPIS2 find_VV0 the_AT similar_JJ neighboring_JJ component_NN1 pair_NN having_VHG same_DA spatial_JJ distance_NN1 ._. 
</s>
<s>
An_AT1 inapplicable_JJ rule_NN1 is_VBZ simply_RR ignored_VVN ._. 
</s>
<s>
Although_CS the_AT particular_JJ vicissitudes_NN2 of_IO ContactPoint_NP1 are_VBR not_XX directly_RR addressed_VVN ,_, it_PPH1 is_VBZ pertinent_JJ to_TO ask_VVI why_RRQ it_PPH1 drew_VVD so_RG little_DA1 attention_NN1 from_II IS_VBZ scholars_NN2 ._. 
</s>
<s>
The_AT provisional_JJ interview_NN1 script_NN1 was_VBDZ pre-tested_JJ with_IW fellow_JJ academics_NN2 to_TO ensure_VVI its_APPGE clarity_NN1 and_CC relevance_NN1 ,_, and_CC then_RT pilot_NN1 tested_VVN with_IW five_MC of_IO the_AT original_JJ respondents_NN2 who_PNQS agreed_VVD to_TO participate_VVI in_II the_AT qualitative_JJ phase_NN1 ._. 
</s>
<s>
It_PPH1 ensures_VVZ that_CST each_DD1 participating_JJ network_NN1 (_( autonomous_JJ system_NN1 )_) has_VHZ a_AT1 route_NN1 for_IF reaching_VVG every_AT1 block_NN1 of_IO IP_NP1 addresses_VVZ ,_, known_VVN as_II prefixes_NN2 ._. 
</s>
<s>
Identity_NN1 is_VBZ the_AT primary_JJ aspect_NN1 of_IO privacy_NN1 and_CC refers_VVZ to_II hiding_VVG or_CC masking_NN1 the_AT identity_NN1 of_IO involved_JJ persons_NN2 ._. 
</s>
<s>
For_IF items_NN2 1_MC1 ,_, 3_MC ,_, 5_MC ,_, 7_MC ,_, and_CC 9_MC the_AT score_NN1 contribution_NN1 is_VBZ the_AT scale_NN1 position_NN1 minus_II 1_MC1 ._. 
</s>
<s>
Strong_JJ coupling_NN1 is_VBZ needed_VVN between_II these_DD2 systems_NN2 to_TO cope_VVI with_IW the_AT re-versioning_NN1 and_CC reuse_NN1 of_IO documents_NN2 ,_, the_AT evolution_NN1 of_IO the_AT ontologies_NN2 used_VMK to_TO describe_VVI them_PPHO2 and_CC a_AT1 range_NN1 of_IO different_JJ users_NN2 who_PNQS may_VM require_VVI different_JJ views_NN2 on_II the_AT data_NN or_CC have_VH0 different_JJ access_NN1 rights_NN2 ._. 
</s>
<s>
Now_RT we_PPIS2 can_VM use_VVI the_AT cost_NN1 model_NN1 for_IF both_DB2 mapping_VVG results_NN2 ._. 
</s>
<s>
This_DD1 is_VBZ because_CS ,_, the_AT radiative_JJ losses_NN2 predominantly_RR impact_VV0 only_RR the_AT regions_NN2 near_II the_AT x=0cm_FO and_CC x=1_FO ,_, cm_NNU boundaries_NN2 than_CSN the_AT central_JJ region_NN1 of_IO the_AT domain_NN1 ._. 
</s>
<s>
We_PPIS2 implement_VV0 most_DAT (_( although_CS not_XX all_RR )_) of_IO the_AT optimizations_NN2 we_PPIS2 have_VH0 surveyed_VVN ._. 
</s>
<s>
Moreover_RR ,_, we_PPIS2 intend_VV0 to_TO improve_VVI this_DD1 approach_NN1 so_CS21 that_CS22 it_PPH1 can_VM discover_VVI shared_JJ services_NN2 that_CST are_VBR not_XX obvious_JJ to_II superficial_JJ human-based_JJ examination_NN1 ._. 
</s>
<s>
Although_CS many_DA2 readers_NN2 might_VM not_XX consider_VVI 34%_NNU to_TO represent_VVI a_AT1 universal_JJ problem_NN1 ,_, it_PPH1 is_VBZ large_JJ enough_RR to_TO justify_VVI concern_NN1 and_CC indicate_VVI a_AT1 need_NN1 for_IF more_DAR studies_NN2 and_CC more_DAR care_NN1 when_CS designing_VVG and_CC implementing_VVG security_NN1 controls_NN2 ._. 
</s>
<s>
This_DD1 type_NN1 of_IO reliability_NN1 could_VM be_VBI applied_VVN to_II the_AT repertory_NN1 grid_NN1 technique_NN1 since_CS one_PN1 would_VM typically_RR expect_VVI to_TO significant_JJ correlations_NN2 among_II constructs_NN2 within_II a_AT1 grid_NN1 ,_, suggesting_VVG that_CST they_PPHS2 are_VBR related_VVN in_II focus_NN1 ._. 
</s>
<s>
To_TO rephrase_VVI a_AT1 point_NN1 from_II our_APPGE earlier_JJR work_NN1 (_( Ye_PPY and_CC Smith_NP1 ,_, 2002_MC )_) ,_, the_AT community_NN1 insists_VVZ on_II strict_JJ access_NN1 controls_NN2 protecting_VVG the_AT client_NN1 file_NN1 system_NN1 from_II server_NN1 content_NN1 ,_, but_CCB neglects_VVZ access_NN1 controls_NN2 protecting_VVG the_AT user_NN1 '_NULL s_ZZ1 perception_NN1 of_IO the_AT client_NN1 user_NN1 interface_NN1 ._. 
</s>
<s>
Content_JJ units_NN2 like_II wedding_NN1 can_VM have_VHI very_RG high_JJ frequency_NN1 owing_II21 to_II22 the_AT popularity_NN1 of_IO the_AT event_NN1 or_CC concept_NN1 ;_; however_RR ,_, co-occurrence_NN1 statistics_NN help_VV0 push_VVI such_DA candidates_NN2 lower_VV0 down_RP the_AT list_NN1 (_( from_II Rank_NP1 138_MC in_II frequency_NN1 to_II out_II21 of_II22 the_AT top-200_MC by_II all_DB other_JJ indicators_NN2 )_) ._. 
</s>
<s>
Many_DA2 of_IO them_PPHO2 have_VH0 decision_NN1 making_VVG power_NN1 in_II the_AT community_NN1 and_CC understood_VVD the_AT relative_JJ feasibility_NN1 and_CC cost-effectiveness_NN1 of_IO proposed_JJ solutions_NN2 ._. 
</s>
<s>
Second_NNT1 the_AT projections_NN2 of_IO the_AT model_NN1 were_VBDR compared_VVN with_IW data_NN from_II real-world_JJ projects_NN2 (_( criteria_NN2 9_MC and_CC 10_MC of_IO the_AT quality_NN1 framework_NN1 )_) ._. 
</s>
<s>
Letjforj=-N+1_FO ,_, ,_, Nbe_VV0 the_AT index_NN1 of_IO thej_JJ th_NNU repatom_NN1 ,_, where2Nis_FO the_AT number_NN1 of_IO repatoms_NN2 ._. 
</s>
<s>
However_RR ,_, only_RR two_MC studies_NN2 (_( discussed_VVN below_RL )_) were_VBDR found_VVN that_CST focused_VVD on_II analyzing_VVG personality_NN1 from_II text_NN1 ._. 
</s>
<s>
The_AT finite_JJ element_NN1 node_NN1 at_II (_( L_ZZ1 ,_, 0_MC )_) is_VBZ duplicated_VVN so_CS21 that_CS22 the_AT exact_JJ essential_JJ boundary_NN1 condition_NN1 can_VM be_VBI realized_VVN ._. 
</s>
<s>
Measuring_VVG Knowledge_NN1 Diffusion_NN1 (_( via_II SNA_NN1 metrics_NN2 )_) :_: Social_JJ Network_NN1 Analysis_NN1 (_( SNA_NN1 )_) is_VBZ used_VVN to_TO quantify_VVI aspects_NN2 of_IO network_NN1 structures_NN2 in_BCL21 order_BCL22 to_TO support_VVI pattern_NN1 identification_NN1 in_II communication_NN1 networks_NN2 &lsqb;_( 60_MC &rsqb;_) ._. 
</s>
<s>
However_RR ,_, we_PPIS2 have_VH0 implemented_VVN a_AT1 simplified_JJ version_NN1 of_IO the_AT referenced_JJ metamodel_NN1 in_BCL21 order_BCL22 to_TO exclude_VVI different_JJ types_NN2 of_IO containment_NN1 relations_NN2 and_CC attribute_VV0 definitions_NN2 that_CST are_VBR out_II21 of_II22 our_APPGE problem_NN1 scope_NN1 ._. 
</s>
<s>
To_TO test_VVI the_AT performance_NN1 of_IO our_APPGE DE_NP1 schemes_NN2 when_RRQ the_AT size_NN1 of_IO the_AT problems_NN2 is_VBZ increased_VVN (_( Section_NN1 6.2_MC )_) ._. 
</s>
<s>
Models_NN2 become_VV0 increasingly_RR hard_JJ to_TO analyze_VVI ,_, understand_VVI ,_, and_CC interpret_VVI ._. 
</s>
<s>
The_AT overhead_JJ stems_NN2 from_II the_AT fact_NN1 that_CST a_AT1 real-_NN1 and_CC a_AT1 hyper-thread_NN1 share_VV0 the_AT floating_JJ point_NN1 unit_NN1 ._. 
</s>
<s>
Our_APPGE own_DA created_JJ English_JJ language_NN1 SGGS-2_MC database_NN1 consists_VVZ of_IO 360_MC sentences_NN2 ,_, 10_MC sentences_NN2 spoken_VVN by_II each_DD1 of_IO 36_MC (_( all_DB males_NN2 )_) speakers_NN2 obtained_VVN in_II two_MC different_JJ sessions_NNT2 within_II a_AT1 span_NN1 of_IO 2_MC months_NNT2 ._. 
</s>
<s>
In_II all_DB cases_NN2 K_ZZ1 is_VBZ set_VVN to_II 25_MC ._. 
</s>
<s>
For_REX21 example_REX22 ,_, it_PPH1 could_VM become_VVI a_AT1 standard_NN1 in_II publications_NN2 to_TO also_RR provide_VVI diagnostic_JJ and_CC evaluation_NN1 indicators_NN2 that_CST put_VV0 a_AT1 single_JJ model_NN1 in_II31 relation_II32 to_II33 other_JJ IAMs_NP2 ._. 
</s>
<s>
It_PPH1 should_VM be_VBI chosen_VVN accordingly_RR to_II the_AT richness_NN1 and_CC complexity_NN1 of_IO the_AT dataset_NN1 ._. 
</s>
<s>
Accordingly_RR ,_, the_AT broker_NN1 references_NN2 included_VVN in_II the_AT service_NN1 descriptor_NN1 will_VM instruct_VVI the_AT data_NN source_NN1 administrator_NN1 of_IO the_AT brokers_NN2 to_II which_DDQ they_PPHS2 must_VM publish_VVI the_AT deployed_JJ services_NN2 ._. 
</s>
<s>
We_PPIS2 observe_VV0 a_AT1 range_NN1 of_IO between_II 15.6%_FO (_( for_IF P6_FO )_) and_CC 41.6%_FO (_( for_IF P10_FO )_) for_IF Others_NN2 '_NULL use_NN1 of_IO social_JJ and_CC positive_JJ words_NN2 ,_, and_CC between_II 18.9%_FO (_( for_IF P6_FO )_) and_CC 23.8%_FO (_( for_IF P3_FO )_) for_IF complex_JJ word_NN1 usage_NN1 respectively_RR ._. 
</s>
<s>
While_CS the_AT first_MD two_MC challenges_NN2 are_VBR applicable_JJ for_IF all_DB ABMs_NP2 ,_, this_DD1 one_PN1 is_VBZ inherited_VVN by_II spatial_JJ ABMs_NP2 from_II the_AT necessity_NN1 to_TO have_VHI a_AT1 landscape_NN1 ._. 
</s>
<s>
Decompose_VV0 the_AT velocity_NN1 field_NN1 inside_II the_AT swimming_NN1 body_NN1 as_CSA u=ur+uf_FO ,_, where_CS ur_FO is_VBZ the_AT rigid_JJ motion_NN1 component_NN1 of_IO the_AT body_NN1 velocity_NN1 and_CC uf_NN1 is_VBZ the_AT deforming_JJ motion_NN1 component_NN1 of_IO the_AT body_NN1 velocity_NN1 in_II its_APPGE own_DA reference_NN1 frame_NN1 ._. 
</s>
<s>
Then_RT ,_, the_AT victim_NN1 AS_CSA announces_VVZ its_APPGE address_NN1 blocks_NN2 to_II prime_JJ the_AT history-based_JJ registry_NN1 of_IO each_DD1 PGBGP-enabled_JJ router_NN1 ._. 
</s>
<s>
The_AT reservation_NN1 fails_VVZ only_RR when_CS the_AT sum_NN1 of_IO the_AT times_NNT2 spent_VVD waiting_VVG at_II all_DB hops_NN2 exceeds_VVZ the_AT defer_VV0 bound_VVN ._. 
</s>
<s>
In_II the_AT case_NN1 wherek=1-1_FO ,_, and_CC especially_RR when1r-C1_JJ ,_, a_AT1 choice_NN1 of_IO a_AT1 much_RR shorter_JJR long_JJ burst_JJ boundary_NN1 might_VM produce_VVI a_AT1 much_RR more_RGR likely_JJ event_NN1 ._. 
</s>
<s>
Magpie_NP1 &lsqb;_( 55_MC &rsqb;_) ,_, for_REX21 example_REX22 operates_VVZ from_II within_II a_AT1 web_NN1 browser_NN1 and_CC does_VDZ "_" real-time_JJ "_" annotation_NN1 of_IO web_NN1 resources_NN2 by_II highlighting_VVG text_NN1 strings_NN2 related_VVN to_II an_AT1 ontology_NN1 of_IO the_AT user_NN1 '_NULL s_ZZ1 choice_NN1 ._. 
</s>
<s>
Thus_RR Coulomb_VV0 '_NULL s_ZZ1 law_NN1 suffices_VVZ to_TO explain_VVI covalent_JJ and_CC noncovalent_JJ bonding_NN1 ._. 
</s>
<s>
Users_NN2 were_VBDR given_VVN four_MC options_NN2 :_: (_( i_ZZ1 )_) based_VVN on_II the_AT given_JJ knowledge_NN1 ,_, I_PPIS1 am_VBM confident_JJ that_CST the_AT conclusion_NN1 is_VBZ correct_JJ ;_; (_( ii_MC )_) based_VVN on_II the_AT given_JJ knowledge_NN1 ,_, I_PPIS1 think_VV0 the_AT conclusion_NN1 is_VBZ more_RGR or_CC less_RGR plausible_JJ ;_; (_( iii_MC )_) the_AT given_JJ knowledge_NN1 does_VDZ not_XX support_VVI the_AT conclusion_NN1 ;_; (_( iv_MC )_) I_PPIS1 don_VV0 '_NULL t_ZZ1 know_VV0 ._. 
</s>
<s>
This_DD1 is_VBZ more_RGR likely_JJ when_CS the_AT parts_NN2 are_VBR small_JJ ,_, simple_JJ ,_, or_CC both_DB2 ._. 
</s>
<s>
Resilience_NN1 evaluation_NN1 is_VBZ more_RGR difficult_JJ than_CSN evaluating_VVG networks_NN2 in_II31 terms_II32 of_II33 traditional_JJ security_NN1 metrics_NN2 ,_, due_II21 to_II22 the_AT need_NN1 to_TO evaluate_VVI the_AT ability_NN1 of_IO the_AT network_NN1 to_TO continue_VVI providing_VVG an_AT1 acceptable_JJ level_NN1 of_IO service_NN1 ,_, while_CS withstanding_VVG challenges_NN2 as_CSA defined_VVN in_II Section_NN1 2.3.1_MC &lsqb;_( 167_MC &rsqb;_) and&lsqb;168&rsqb;_FO ._. 
</s>
<s>
The_AT integration_NN1 using_VVG 5×5_FO tensor-product_JJ rule_NN1 yields_VVZ a_AT1 rate_NN1 of_IO convergence_NN1 of_IO 0.85_MC and_CC 0.76_MC for_IF enrichment_NN1 radii_NN2 of_IO re=0.25_FO and_CC re=0.5_FO ,_, respectively_RR ,_, whereas_CS more_RGR accurate_JJ integration_NN1 with_IW the_AT 5×5_FO generalized_JJ Duffy_NP1 transformation_NN1 is_VBZ realized_VVN ,_, which_DDQ delivers_VVZ the_AT optimal_JJ first-order_JJ rate_NN1 of_IO convergence_NN1 in_II the_AT energy_NN1 norm_NN1 ._. 
</s>
<s>
In_II our_APPGE current_JJ system_NN1 ,_, we_PPIS2 have_VH0 implemented_VVN a_AT1 popular_JJ graph-based_JJ technique_NN1 to_TO deal_VVI with_IW ontologiessimilarity_NN1 propagation_NN1 in_II the_AT structure_NN1 level_NN1 matcher_NN1 module_NN1 (_( see_VV0 Section_NN1 6for_FO more_RRR details_NN2 )_) ._. 
</s>
<s>
As_CSA described_VVN in_II Section4.4_FO ,_, one_MC1 of_IO the_AT recommendation_NN1 approaches_NN2 presented_VVN in_II this_DD1 paper_NN1 ,_, the_AT global_JJ recommendation_NN1 ,_, is_VBZ an_AT1 extension_NN1 of_IO a_AT1 recommendation_NN1 approach_NN1 developed_VVD previously_RR (_( Hopfgartner_NP1 et_RA21 al._RA22 ,_, 2008_MC )_) ._. 
</s>
<s>
The_AT Khachiyan_JJ algorithm_NN1 finds_VVZ the_AT minimum_JJ volume_NN1 ellipsoid_NN1 (_( characterized_VVN by_II the_AT matrix_NN1 A_ZZ1 and_CC its_APPGE center_NN1 coordinates_NN2 as_CSA shown_VVN in_II (_( 1_MC1 )_) )_) by_II an_AT1 iterative_JJ procedure_NN1 constructing_VVG a_AT1 sequence_NN1 of_IO decreasing_JJ ellipsoids_NN2 ._. 
</s>
<s>
The_AT reliability_NN1 assessment_NN1 of_IO our_APPGE quality_NN1 evaluation_NN1 was_VBDZ based_VVN on_II 54_MC papers_NN2 ._. 
</s>
<s>
It_PPH1 is_VBZ important_JJ to_TO note_VVI that_CST we_PPIS2 are_VBR not_XX advocating_VVG the_AT complete_JJ or_CC exclusive_JJ use_NN1 of_IO ontologies_NN2 throughout_II the_AT entire_JJ system_NN1 ._. 
</s>
<s>
The_AT measure_NN1 False_JJ Positive_JJ Rate_NN1 (_( FPR_NP1 )_) relates_VVZ the_AT files_NN2 incorrectly_RR classified_VVN as_CSA failure-prone_JJ to_II the_AT total_JJ number_NN1 of_IO non-failure-prone_JJ files_NN2 ._. 
</s>
<s>
Moreover_RR ,_, better_RRR training-issue_JJ filtering_NN1 algorithms_NN2 need_VV0 to_TO be_VBI implemented_VVN in_BCL21 order_BCL22 to_TO make_VVI better_JJR use_NN1 ofLeast_NN1 Squares_NN2 Learningin_NN1 computing_NN1 weights_NN2 ._. 
</s>
<s>
Currently_RR this_DD1 is_VBZ awaiting_VVG industrial_JJ commitment_NN1 ,_, since_CS only_JJ telephony_NN1 companies_NN2 have_VH0 access_NN1 to_II a_AT1 representative_JJ user_NN1 base_NN1 ._. 
</s>
<s>
In_II this_DD1 case_NN1 ,_, once_CS the_AT rules_NN2 have_VH0 been_VBN formalized_VVN ,_, the_AT rule_NN1 engine_NN1 can_VM run_VVI them_PPHO2 without_IW further_JJR programming_NN1 effort_NN1 ._. 
</s>
<s>
Our_APPGE DNS_NN2 tunnel_VV0 detection_NN1 results_NN2 are_VBR shown_VVN in_II Table_NN1 8_MC ._. 
</s>
<s>
Therefore_RR non-equivalence_JJ is_VBZ in_RP ,_, that_DD1 is_VBZ and_CC equivalence_NN1 is_VBZ in_RP ._. 
</s>
<s>
Weighting_VVG in_II31 favor_II32 of_II33 the_AT defer_VV0 time_NNT1 can_VM cause_VVI larger_JJR requests_NN2 to_TO be_VBI dropped_VVN since_CS the_AT priority_NN1 is_VBZ the_AT order_NN1 of_IO arrival_NN1 ._. 
</s>
<s>
Let_VV0 A_ZZ1 be_VBI the_AT pre-model_NN1 obtained_VVN in_II line_NN1 3_MC ._. 
</s>
<s>
The_AT main_JJ practical_JJ problem_NN1 is_VBZ that_CST it_PPH1 does_VDZ not_XX include_VVI any_DD field_NN1 categorisations_NN2 as_CSA does_VDZ WoS_NP2 ._. 
</s>
<s>
It_PPH1 might_VM be_VBI a_AT1 valuable_JJ activity_NN1 for_IF scholarly_JJ associations_NN2 to_TO perhaps_RR agree_VVI lists_NN2 of_IO journals_NN2 that_CST are_VBR relevant_JJ to_II their_APPGE disciplines_NN2 without_IW ,_, of_RR21 course_RR22 ,_, assessing_VVG their_APPGE quality_NN1 ._. 
</s>
<s>
In_II contrast_NN1 ,_, search/index_NN1 and_CC index/index_VV0 redundancies_NN2 involve_VV0 unnecessary_JJ work_NN1 being_VBG done_VDN at_II two_MC different_JJ peers_NN2 ,_, and_CC it_PPH1 is_VBZ difficult_JJ for_IF those_DD2 peers_NN2 to_TO co-ordinate_VVI and_CC discover_VVI that_CST their_APPGE work_NN1 is_VBZ redundant_JJ ._. 
</s>
<s>
For_REX21 instance_REX22 ,_, school_NN1 A_ZZ1 had_VHD continuously_RR rising_VVG citation_NN1 rates_NN2 but_CCB when_CS normalised_VVN they_PPHS2 actually_RR fell_VVD over_RP some_DD years_NNT2 because_CS the_AT citations_NN2 in_II the_AT field_NN1 as_II a_AT1 whole_NN1 were_VBDR rising_VVG ._. 
</s>
<s>
The_AT current_JJ GENI_JJ deployment_NN1 is_VBZ actively_RR growing_VVG to_II over_RG fifty_MC sites_NN2 in_II the_AT current_JJ expansion_NN1 phase_NN1 and_CC is_VBZ expected_VVN to_TO continue_VVI growing_VVG to_II an_AT1 at_II scale_NN1 size_NN1 of_IO one_MC1 hundred_NNO to_II two_MC hundred_NNO sites_NN2 ._. 
</s>
<s>
While_CS the_AT behaviour_NN1 of_IO MI_NP1 (_( T_ZZ1 ,_, Ps_NP1 ,_, Pt_FO )_) from_II a_AT1 general_JJ perspective_NN1 is_VBZ of_IO some_DD interest_NN1 ,_, The_AT following_JJ theorem_NN1 is_VBZ a_AT1 direct_JJ consequence_NN1 of_IO the_AT submodularity_NN1 of_IO ?_? ?_? t(T)_UH and_CC the_AT approximation_NN1 result_NN1 of_IO ?_? &lsqb;_( 39_MC &rsqb;_) ._. 
</s>
<s>
The_AT Emissions_NN2 Trading_NN1 System_NN1 concept_NN1 defines_VVZ the_AT ETS_NP2 scope_NN1 through_II the_AT concepts_NN2 strictly_RR referring_VVG to_II the_AT EU_NP1 directives_NN2 (_( Grubb_NP1 and_CC Neuhoff_NP1 ,_, 2006_MC ,_, Ellerman_NP1 and_CC Buchner_NP1 ,_, 2007_MC )_) (_( e.g._REX ,_, non_FU CO2_FO gases_NN2 are_VBR outside_II this_DD1 specific_JJ scope_NN1 even_CS21 if_CS22 they_PPHS2 are_VBR in_II the_AT scope_NN1 of_IO the_AT whole_JJ ontology_NN1 )_) ._. 
</s>
<s>
However_RR ,_, we_PPIS2 believe_VV0 the_AT distinction_NN1 between_II objects_NN2 and_CC domain_NN1 values_NN2 is_VBZ important_JJ since_CS not_XX every_AT1 complex-domain_JJ structure_NN1 also_RR meets_VVZ the_AT intuitive_JJ concept_NN1 of_IO an_AT1 object_NN1 ._. 
</s>
<s>
A_AT1 point_NN1 to_TO be_VBI noted_VVN is_VBZ that_DD1 batch_NN1 rekeying_NN1 provides_VVZ a_AT1 trade-off_NN1 between_II performance_NN1 and_CC security_NN1 ._. 
</s>
<s>
In_II the_AT next_MD section_NN1 ,_, I_PPIS1 draw_VV0 from_II the_AT foundational_JJ literature_NN1 to_TO discuss_VVI innovating_JJ with_IW IT_NN1 among_II firms_NN2 ,_, setting_VVG the_AT stage_NN1 ._. 
</s>
<s>
In_II this_DD1 section_NN1 we_PPIS2 present_VV0 the_AT results_NN2 of_IO our_APPGE case_NN1 study_NN1 ._. 
</s>
<s>
Since_CS in_II our_APPGE evaluation_NN1 we_PPIS2 determine_VV0 the_AT size_NN1 of_IO a_AT1 graph_NN1 chunk_NN1 by_II the_AT number_NN1 of_IO triples_NN2 ,_, the_AT imbalanced_JJ chunk_NN1 sizes_NN2 are_VBR caused_VVN by_II different_JJ number_NN1 of_IO incident_NN1 edges_NN2 that_CST are_VBR assigned_VVN to_II the_AT different_JJ graph_NN1 chunks_NN2 ._. 
</s>
<s>
Hence_RR ,_, using_VVG the_AT criterion_NN1 described_VVN in_II Section_NN1 2_MC ,_, we_PPIS2 have_VH0 not_XX identified_VVN any_DD frequently_RR cited_VVN (_( >50%_FO )_) ,_, '_NULL project_NN1 planning_NN1 '_NULL practice_NN1 which_DDQ has_VHZ a_AT1 '_NULL high_JJ '_NULL perceived_JJ value_NN1 ._. 
</s>
<s>
As_CSA we_PPIS2 demonstrate_VV0 in_II Table_NN1 2_MC ,_, the_AT summation_NN1 classifier_NN1 can_VM also_RR closely_RR approximate_VVI both_DB2 readers_NN2 ,_, the_AT minimum_NN1 and_CC maximum_JJ classifiers_NN2 ._. 
</s>
<s>
For_IF practitioners_NN2 ,_, this_DD1 research_NN1 presents_VVZ specific_JJ examples_NN2 of_IO ways_NN2 that_CST innovative_JJ CIO_NN1 '_NULL s_ZZ1 work_VV0 ,_, both_RR interpersonally_RR and_CC institutionally_RR ._. 
</s>
<s>
Composite_JJ change_NN1 detection_NN1 :_: WebVigiL_NP1 provides_VVZ an_AT1 elegant_JJ way_NN1 to_TO specify_VVI multiple_JJ change_NN1 types_NN2 ,_, such_II21 as_II22 "_" changes_NN2 occurring_VVG to_II either_RR images_NN2 or_CC links_NN2 "_" ,_, which_DDQ none_PN of_IO the_AT above_JJ systems_NN2 provide_VV0 ._. 
</s>
<s>
Even_RR with_IW an_AT1 accurate_JJ initialization_NN1 (_( which_DDQ can_VM be_VBI a_AT1 rough_JJ binarization_NN1 map_NN1 )_) ,_, the_AT final_JJ factor_NN1 that_CST determines_VVZ the_AT output_NN1 is_VBZ the_AT complexity_NN1 of_IO the_AT Markov_NP1 model_NN1 which_DDQ is_VBZ lower_JJR than_CSN the_AT complexity_NN1 of_IO the_AT degradation_NN1 ._. 
</s>
<s>
From_II Theorem_NN1 3.1_MC it_PPH1 follows_VVZ that_CST for_IF formulas_NN2 in_RP ,_, we_PPIS2 can_VM get_VVI the_AT same_DA axiomatization_NN1 with_II31 respect_II32 to_II33 structures_NN2 in_II M()_NN1 for_IF both_DB2 the_AT out_RP and_CC in_II semantics_NN1 ;_; moreover_RR ,_, this_DD1 axiomatization_NN1 is_VBZ the_AT same_DA as_CSA that_DD1 for_IF the_AT common-interpretation_JJ case_NN1 ._. 
</s>
<s>
Since_CS a_AT1 HRP_NP1 is_VBZ involved_JJ in_II a_AT1 high-risk_JJ cycle_NN1 where_CS crime_NN1 is_VBZ linked_VVN to_II a_AT1 drug_NN1 habit_NN1 ,_, there_EX is_VBZ a_AT1 probability_NN1 associated_VVN with_IW them_PPHO2 being_VBG sent_VVN to_II incapacitation_NN1 ._. 
</s>
<s>
However_RR ,_, for_IF irregular_JJ sensor_NN1 fields_NN2 ,_, part_NN1 of_IO the_AT data_NN dissemination_NN1 circle_NN1 might_VM fall_VVI outside_RL ,_, and_CC the_AT aggregation_NN1 point_NN1 in_II the_AT double_JJ rulings_NN2 model_NN1 will_VM be_VBI deteriorated_VVN into_II two_MC or_CC more_RGR touching_JJ points_NN2 between_II the_AT dissemination_NN1 circle_NN1 and_CC the_AT field_NN1 boundary_NN1 ._. 
</s>
<s>
Although_CS we_PPIS2 identified_VVD broadly_RR which_DDQ questions_NN2 would_VM be_VBI inappropriate_JJ for_IF certain_JJ types_NN2 of_IO study_NN1 ,_, we_PPIS2 found_VVD some_DD questions_NN2 were_VBDR inappropriate_JJ due_II21 to_II22 the_AT context_NN1 of_IO the_AT study_NN1 ._. 
</s>
<s>
More_RGR detailed_JJ information_NN1 on_II land_NN1 cover_NN1 was_VBDZ further_RRR obtained_VVN from_II the_AT 2000_MC General_JJ Agriculture_NN1 Census_NN1 for_IF communes_NN2 and_CC from_II classified_JJ Landsat_NP1 sensed_VVD satellite_NN1 data_NN (_( Landsat_NP1 5_MC ?_? 
</s>
<s>
TM_NP1 )_) ._. 
</s>
<s>
This_DD1 was_VBDZ greatly_RR helped_VVN by_II the_AT fact_NN1 that_CST the_AT interviewer_NN1 was_VBDZ a_AT1 summer_NNT1 intern_NN1 and_CC was_VBDZ therefore_RR new_JJ and_CC unknown_JJ to_II the_AT other_JJ employees_NN2 ._. 
</s>
<s>
In_II Fig._NN1 2(b)_FO we_PPIS2 show_VV0 a_AT1 class_NN1 of_IO oval-shaped_JJ seal_NN1 having_VHG noise_NN1 (_( missing_JJ seal_NN1 information_NN1 or_CC presence_NN1 of_IO other_JJ text_NN1 characters_NN2 from_II document_NN1 )_) ._. 
</s>
<s>
So_RG far_RR ,_, unfortunately_RR ,_, there_EX are_VBR very_RG few_DA2 examples_NN2 of_IO frameworks_NN2 being_VBG used_VVN beyond_II their_APPGE respective_JJ development_NN1 teams_NN2 ._. 
</s>
<s>
Again_RT ,_, the_AT axes_NN2 shown_VVN are_VBR those_DD2 used_VMK to_TO construct_VVI the_AT rectangle_NN1 defining_VVG the_AT scenario_NN1 and_CC result_VV0 from_II the_AT two_MC input_NN1 dimensions_NN2 PRIM_JJ and_CC CPCA-PRIM_JJ have_VH0 identified_VVN as_II the_AT most_RGT influential_JJ ._. 
</s>
<s>
In_II OWL_NN1 DL_MC all_DB classes_NN2 are_VBR interpreted_VVN as_CSA subsets_NN2 of_IO the_AT abstract_JJ domain_NN1 ,_, and_CC for_IF each_DD1 constructor_NN1 the_AT semantics_NN1 of_IO the_AT resulting_JJ class_NN1 is_VBZ defined_VVN in_II31 terms_II32 of_II33 the_AT semantics_NN1 of_IO its_APPGE components_NN2 ._. 
</s>
<s>
Section_NN1 2.1_MC presents_VVZ the_AT background_NN1 on_II PS_NN1 and_CC PG_NP1 ._. 
</s>
<s>
The_AT answer_NN1 of_IO the_AT user_NN1 in_II each_DD1 question_NN1 is_VBZ binary_JJ ,_, either_RR correct_JJ or_CC incorrect_JJ ._. 
</s>
<s>
Most_DAT of_IO the_AT existing_JJ literature_NN1 studies_NN2 the_AT two_MC test_NN1 selection_NN1 methods_NN2 disjointly_RR ._. 
</s>
<s>
The_AT large_JJ number_NN1 of_IO intermediate_JJ months_NNT2 during_II which_DDQ the_AT emissions_NN2 volumes_NN2 are_VBR significantly_RR higher_JJR than_CSN those_DD2 required_VVN to_TO reach_VVI the_AT final_NN1 L=900_FO limit_NN1 suggests_VVZ that_CST ,_, in_II some_DD respects_NN2 ,_, an_AT1 instantaneous_JJ cap_NN1 reduction_NN1 may_VM be_VBI a_AT1 more_RGR efficient_JJ way_NN1 of_IO reducing_VVG carbon_NN1 emissions_NN2 than_CSN a_AT1 gradual_JJ change_NN1 ._. 
</s>
<s>
These_DD2 algorithms_NN2 both_RR take_VV0 a_AT1 changeset_NN1 –_- containing_VVG additions_NN2 and_CC deletions_NN2 –_- asinput_VV0 ,_, and_CC append_VV0 it_PPH1 as_II a_AT1 new_JJ version_NN1 to_II the_AT store_NN1 ._. 
</s>
<s>
Moreover_RR ,_, we_PPIS2 develop_VV0 a_AT1 conjugate_NN1 prior_RR for_IF the_AT Beta-Liouville_NP1 distribution_NN1 taking_VVG into_II account_NN1 the_AT fact_NN1 that_CST it_PPH1 belongs_VVZ to_II the_AT exponential_NN1 family_NN1 of_IO distributions_NN2 ._. 
</s>
<s>
We_PPIS2 adopt_VV0 Tversky_NP1 '_NULL s_ZZ1 interpretation_NN1 of_IO similarity_NN1 ,_, and_CC thus_RR seek_VV0 to_TO express_VVI these_DD2 operations_NN2 T_ZZ1 and_CC T_ZZ1 in_II some_DD representation_NN1 which_DDQ both_RR is_VBZ robust_JJ and_CC affords_VVZ sufficient_JJ feature_NN1 production_NN1 to_TO permit_VVI feature_NN1 matching_NN1 &lsqb;_( 4_MC &rsqb;_) ._. 
</s>
<s>
ApEn_NP1 differs_VVZ from_II SampEn_NP1 in_II that_DD1 its_APPGE calculation_NN1 involves_VVZ counting_VVG a_AT1 self-match_NN1 for_IF each_DD1 sequence_NN1 of_IO a_AT1 pattern_NN1 ,_, which_DDQ leads_VVZ to_TO bias_VVI in_II ApEn_NP1 &lsqb;_( 11_MC &rsqb;_) and_CC &lsqb;_( 12_MC &rsqb;_) ._. 
</s>
<s>
According_CS the_AT European_JJ Commission_NN1 in_II 2011_MC ,_, there_EX were_VBDR two_MC main_JJ principles_NN2 which_DDQ defined_VVD the_AT Core_NN1 vocabularies_NN2 &lsqb;_( 30_MC &rsqb;_) :_: (_( i_ZZ1 )_) highly_RR reusable_JJ :_: the_AT specification_NN1 is_VBZ simple_JJ and_CC captures_VVZ basic_JJ and_CC generic_JJ characteristics_NN2 of_IO an_AT1 information_NN1 entity_NN1 ,_, regardless_RR of_IO the_AT context_NN1 this_DD1 entity_NN1 is_VBZ used_VVN and_CC (_( ii_MC )_) extensible_JJ :_: domain_NN1 specific_JJ specialisations_NN2 can_VM be_VBI drafted_VVN on_II31 top_II32 of_II33 the_AT core_NN1 representation_NN1 ._. 
</s>
<s>
There_EX are_VBR two_MC main_JJ types_NN2 of_IO shape_NN1 description_NN1 methods_NN2 :_: boundary-based_JJ methods_NN2 and_CC region-based_JJ methods_NN2 ._. 
</s>
<s>
Change_NN1 is_VBZ a_AT1 process_NN1 involving_VVG progression_NN1 through_II six_MC stages_NN2 :_: precontemplation_NN1 ,_, contemplation_NN1 (_( thoughts_NN2 )_) ,_, preparation_NN1 (_( thoughts_NN2 and_CC action_NN1 )_) ,_, action_NN1 (_( actual_JJ behavior_NN1 change_NN1 )_) ,_, maintenance_NN1 ,_, and_CC termination_NN1 ._. 
</s>
<s>
The_AT inner_JJ ring_NN1 is_VBZ generated_VVN by_II radar_NN1 echoes_VVZ that_CST experience_VV0 a_AT1 two-bounce_JJ path_NN1 between_II the_AT asphalt_NN1 and_CC the_AT side_NN1 of_IO the_AT vehicle_NN1 ;_; thus_RR ,_, the_AT inner_JJ ring_NN1 represents_VVZ time-of-flight_NN1 to_II the_AT base_NN1 outline_NN1 of_IO the_AT vehicle_NN1 &lsqb;_( 36_MC &rsqb;_) ._. 
</s>
<s>
Therefore_RR modeling_VVG protocols_NN2 should_VM be_VBI simpler_JJR than_CSN continually_RR creating_VVG signatures_NN2 for_IF the_AT latest_JJT malware_NN1 code_NN1 ._. 
</s>
<s>
Application-data_JJ Visible_JJ Entities_NN2 (_( ADV_NP1 )_) :_: ADVs_NP2 are_VBR OpenADN-aware_JJ entities_NN2 that_CST have_VH0 visibility_NN1 (_( and_CC access_NN1 )_) to_II application-level_JJ data_NN ._. 
</s>
<s>
No_AT changes_NN2 were_VBDR performed_VVN to_TO improve_VVI precision_NN1 or_CC recall_VV0 ._. 
</s>
<s>
The_AT LN_NP1 approximations_NN2 were_VBDR more_RGR accurate_JJ than_CSN SW_ND1 solutions_NN2 on_II the_AT level_NN1 four_MC mesh_NN1 ,_, and_CC the_AT CG–V_NN1 velocities_NN2 were_VBDR more_RGR accurate_JJ than_CSN those_DD2 obtained_VVN from_II the_AT CG–S_NP1 approximation_NN1 ,_, apparently_RR due_II21 to_II22 excessive_JJ numerical_JJ diffusion_NN1 from_II the_AT shock_NN1 capturing_VVG term_NN1 ._. 
</s>
<s>
In_II31 addition_II32 to_II33 non-linearity_NN1 ,_, LCP_NP1 is_VBZ more_RGR complex_JJ than_CSN path_NN1 coverage_NN1 problem_NN1 (_( as_CSA defined_VVN in_II PRIDE_NN1 )_) for_IF a_AT1 given_JJ network_NN1 ._. 
</s>
<s>
In_II31 addition_II32 to_II33 the_AT clusters_NN2 ,_, noise_NN1 points_NN2 are_VBR scattered_VVN uniformly_RR in_II the_AT space_NN1 ._. 
</s>
<s>
However_RR ,_, this_DD1 is_VBZ far_RR from_II being_VBG a_AT1 straightforward_JJ task_NN1 ._. 
</s>
<s>
When_CS the_AT device_NN1 processor_NN1 is_VBZ idle_JJ or_CC the_AT device_NN1 is_VBZ being_VBG recharged_VVN ,_, exploiting_VVG periods_NN2 of_IO time_NNT1 when_RRQ retraining_NN1 will_VM have_VHI less_DAR effect_NN1 on_II device_NN1 use_NN1 ._. 
</s>
<s>
The_AT ISA_NP1 Core_NN1 Location_NN1 Vocabulary_NN1 (_( Fig._NN1 8_MC )_) models_VVZ its_APPGE definition_NN1 of_IO "_" Address_NN1 "_" on_II AddressRepresentation_NN1 from_II INSPIRE_VV0 ._. 
</s>
<s>
They_PPHS2 can_VM be_VBI invoked_VVN in_II a_AT1 variety_NN1 of_IO ways_NN2 ,_, facilitating_VVG virtual_JJ applications_NN2 that_CST span_VV0 multiple_JJ organizations_NN2 ,_, data_NN sources_NN2 and_CC computers_NN2 ._. 
</s>
<s>
While_CS powerful_JJ ,_, this_DD1 formalism_NN1 has_VHZ a_AT1 restriction_NN1 that_CST can_VM make_VVI it_PPH1 unsuitable_JJ for_IF modelling_VVG complex_JJ multi-agent_JJ domains_NN2 ._. 
</s>
<s>
Such_DA overview_NN1 provides_VVZ an_AT1 up-to-date_JJ evaluation_NN1 of_IO the_AT way_NN1 in_II which_DDQ ABMs_NP2 have_VH0 been_VBN able_JK to_TO address_VVI the_AT challenges_NN2 and_CC live_VVI up_II21 to_II22 the_AT expectations_NN2 as_II31 well_II32 as_II33 identifying_VVG the_AT remaining_JJ challenges_NN2 ._. 
</s>
<s>
In_RR21 addition_RR22 ,_, the_AT results_NN2 indicate_VV0 that_CST this_DD1 significant_JJ improvement_NN1 in_II outcomes_NN2 does_VDZ not_XX come_VVI at_II the_AT cost_NN1 of_IO longer-term_JJ development_NN1 of_IO the_AT programmers_NN2 '_NULL knowledge_NN1 base_NN1 ,_, in_II this_DD1 case_NN1 their_APPGE higher-level_JJ comprehension_NN1 of_IO the_AT program_NN1 in_II31 relation_II32 to_II33 its_APPGE requirements_NN2 ._. 
</s>
<s>
This_DD1 greedy_JJ algorithm_NN1 also_RR takes_VVZ at_RR21 most_RR22 ?_? 
</s>
<s>
O_ZZ1 (_( m_ZZ1 ?_? log_NN1 ?_? m_ZZ1 ?_? +_FO ?_? 
</s>
<s>
M_ZZ1 )_) time_VV0 to_TO calculate_VVI the_AT minimum_JJ cost_NN1 of_IO routing_VVG through_RP ?_? m_ZZ1 ?_? 
</s>
<s>
MAX_NP1 ISPs_NP2 ._. 
</s>
<s>
The_AT MDS_NP2 algorithm_NN1 requires_VVZ a_AT1 matrix_NN1 of_IO pairwise_RR distances_NN2 between_II samples_NN2 ;_; in_II our_APPGE application_NN1 ,_, we_PPIS2 used_VVD the_AT minimized_JJ LTS-HD_NN1 based_VVN pseudo-distance_NN1 ._. 
</s>
<s>
The_AT reproduction_NN1 of_IO the_AT mean_JJ values_NN2 for_IF each_DD1 of_IO the_AT predictands_NN2 (_( Fig._NN1 @5_FO )_) is_VBZ very_RG accurate_JJ ._. 
</s>
<s>
After_CS classification_NN1 has_VHZ taken_VVN place_NN1 ,_, an_AT1 event_NN1 object_NN1 ,_, as_CSA described_VVN in_II Section_NN1 3.2_MC ,_, is_VBZ created_VVN and_CC added_VVN to_II the_AT buffer_NN1 ._. 
</s>
<s>
The_AT social_JJ and_CC organizational_JJ aspects_NN2 of_IO information_NN1 systems_NN2 development_NN1 are_VBR recognized_VVN ._. 
</s>
<s>
The_AT remainder_NN1 of_IO this_DD1 paper_NN1 explores_VVZ the_AT design_NN1 and_CC practical_JJ use_NN1 of_IO origin_NN1 authentication_NN1 services_NN2 ._. 
</s>
<s>
Usually_RR the_AT data_NN has_VHZ to_TO be_VBI formatted_VVN ,_, sampled_VVD ,_, adapted_VVD ,_, and_CC sometimes_RT transformed_VVN for_IF the_AT data_NN mining_VVG algorithm_NN1 ._. 
</s>
<s>
For_REX21 example_REX22 ,_, to_TO determine_VVI the_AT limit_NN1 equations_NN2 for_IF negligible_JJ fluid_NN1 presence_NN1 ,_, not_XX only_RR must_VM →_NULL 1_MC1 but_CCB f_ZZ1 →_NULL 0_MC also_RR ._. 
</s>
<s>
The_AT results_NN2 showed_VVD that_CST the_AT difference_NN1 between_II versions_NN2 was_VBDZ highly_RR significant_JJ (_( p@<@0.001_FO )_) ._. 
</s>
<s>
For_IF Book_NN1 ,_, hybrids_NN2 with_IW 2-way_JJ perform_VV0 better_RRR than_CSN Pure_JJ Prioritization_NN1 but_CCB are_VBR poor_JJ when_CS compared_VVN to_II Pure_JJ Reduction_NN1 ._. 
</s>
<s>
Taddeo_NN1 argues_VVZ that_CST these_DD2 principles_NN2 are_VBR difficult_JJ to_TO apply_VVI when_RRQ it_PPH1 comes_VVZ to_II cyber_NN1 warfare_NN1 ,_, and_CC that_CST these_DD2 difficulties_NN2 are_VBR worthy_JJ of_IO further_JJR research_NN1 ._. 
</s>
<s>
The_AT PDFs_NP2 of_IO the_AT coefficients_NN2 ,_, as_CSA found_VVN in_II Section_NN1 3.3_MC ,_, are_VBR used_VVN to_TO generate_VVI Monte-Carlo_NP1 samples_NN2 ._. 
</s>
<s>
We_PPIS2 give_VV0 a_AT1 number_NN1 of_IO examples_NN2 of_IO the_AT personalised_JJ feedback_NN1 which_DDQ tuned_VVD appliances_NN2 models_NN2 can_VM be_VBI used_VVN to_TO provide_VVI ._. 
</s>
<s>
All_DB that_DD1 is_VBZ needed_VVN on_II the_AT user_NN1 '_NULL s_ZZ1 part_NN1 is_VBZ to_TO supply_VVI a_AT1 pattern_NN1 that_CST the_AT tool_NN1 will_VM use_VVI to_TO extract_VVI one_MC1 of_IO the_AT identifiers_NN2 that_CST can_VM be_VBI used_VVN for_IF the_AT lookup_NN1 ,_, as_RG well_RR as_CSA to_TO specify_VVI identifier_NN1 '_NULL s_ZZ1 type_NN1 ,_, which_DDQ can_VM be_VBI one_MC1 of_IO the_AT following_JJ :_: name_NN1 ,_, two-letter_JJ code_NN1 (_( ISO_NP1 3166-1_MCMC alpha2_FO )_) and_CC three-letter_JJ code_NN1 (_( ISO_NP1 3166-1_MCMC alpha_NN1 3_MC )_) ._. 
</s>
<s>
For_IF almost_RR all_DB registration_NN1 algorithms_NN2 using_VVG surface_NN1 representation_NN1 ,_, the_AT search_NN1 for_IF closest_RRT points_VVZ is_VBZ required_VVN to_TO obtain_VVI a_AT1 spherical_JJ or_CC cylinder_NN1 regions_NN2 centered_VVD at_II a_AT1 seed_NN1 ._. 
</s>
<s>
Further_JJR research_NN1 is_VBZ suggested_VVN to_TO examine_VVI other_JJ classification_NN1 methods_NN2 for_IF their_APPGE suitability_NN1 to_II the_AT event_NN1 detection_NN1 problem_NN1 ._. 
</s>
<s>
The_AT ODE_NN1 solve_VV0 and_CC RHS_NP1 vector_NN1 assembly_NN1 parts_NN2 of_IO the_AT load_NN1 (_( 20%_NNU and_CC 4%_NNU ,_, respectively_RR of_IO the_AT total_NN1 for_IF the_AT finest_JJT fixed_JJ heart_NN1 mesh_NN1 :_: the_AT two_MC largest_JJT contributions_NN2 after_II the_AT linear_JJ solve_VV0 )_) are_VBR also_RR greatly_RR reduced_VVN (_( speedup_VV0 factors_NN2 of_IO 57_MC and_CC 53_MC ,_, respectively_RR for_IF the_AT heart_NN1 and_CC 89_MC and_CC 105_MC for_IF the_AT slab_NN1 )_) in_II the_AT adaptive_JJ simulations_NN2 as_CSA they_PPHS2 also_RR depend_VV0 on_II the_AT number_NN1 of_IO mesh_NN1 nodes_NN2 ._. 
</s>
<s>
From_II this_DD1 ,_, a_AT1 so-calledmonitoris_NN1 automatically_RR generated_VVN ,_, which_DDQ is_VBZ a_AT1 software_NN1 component_NN1 that_CST ,_, at_II runtime_NNT1 ,_, compares_VVZ an_AT1 observed_JJ behaviour_NN1 of_IO a_AT1 system_NN1 with_IW the_AT specified_JJ reference_NN1 behaviour_NN1 ._. 
</s>
<s>
When_CS we_PPIS2 inspect_VV0 all_DB exit_VV0 violations_NN2 that_CST happen_VV0 in_II the_AT overlay_NN1 link_NN1 AB_FO because_II21 of_II22 a_AT1 transit_NN1 violation_NN1 at_II node_NN1 B_ZZ1 ,_, we_PPIS2 notice_VV0 a_AT1 stronger_JJR correlation_NN1 between_II the_AT AS_CSA experiencing_VVG the_AT most_RRT exit_VV0 violations_NN2 and_CC the_AT node_NN1 B_ZZ1 ,_, rather_CS21 than_CS22 with_IW node_NN1 A._NP1 However_RR ,_, a_AT1 particular_JJ transit_NN1 violating_VVG intermediate_JJ relay_NN1 does_VDZ not_XX have_VHI a_AT1 unique_JJ exit_NN1 violated_VVN AS_CSA preceding_VVG it_PPH1 ,_, i.e._REX there_EX is_VBZ no_AT one-on-one_MC correspondence_NN1 between_II the_AT transit_NN1 violated_VVN AS_CSA and_CC the_AT exit_NN1 violated_VVN AS_II ._. 
</s>
<s>
Wilson_NP1 and_CC Hancock_NP1 have_VH0 previously_RR shown_VVN &lsqb;_( 18_MC &rsqb;_) and_CC &lsqb;_( 19_MC &rsqb;_) how_RRQ permutation_NN1 invariant_JJ polynomials_NN2 can_VM be_VBI used_VVN to_TO derive_VVI features_NN2 which_DDQ describe_VV0 graphs_NN2 and_CC make_VVI full_JJ use_NN1 of_IO the_AT available_JJ spectral_JJ information_NN1 ._. 
</s>
<s>
The_AT set_NN1 of_IO known_JJ event_NN1 classes_NN2 also_RR represents_VVZ the_AT search_NN1 space_NN1 used_VVN for_IF creating_VVG new_JJ hypotheses_NN2 in_II the_AT search_NN1 for_IF rules_NN2 ._. 
</s>
<s>
A_AT1 statistical_JJ assertion_NN1 is_VBZ defined_VVN as_II a_AT1 predicate_NN1 consisting_VVG of_IO two_MC data_NN models_NN2 in_II the_AT form_NN1 of_IO either_DD1 statistical_JJ primitives_NN2 (_( e.g._REX mean_JJ or_CC standard_JJ deviation_NN1 values_NN2 )_) or_CC data_NN models_NN2 (_( e.g._REX histograms_NN2 or_CC density_NN1 functions_NN2 )_) ._. 
</s>
<s>
However_RR ,_, the_AT authors_NN2 have_VH0 not_XX explored_VVN the_AT use_NN1 of_IO this_DD1 specification_NN1 to_TO generate_VVI test_NN1 cases_NN2 in_II TTCN_NP1 by_II using_VVG the_AT MSCs_NP2 as_II test_NN1 purposes_NN2 for_IF an_AT1 SDL_NP1 specification_NN1 ,_, a_AT1 functionality_NN1 supported_VVN by_II Tau_NN1 (_( namely_REX with_IW Autolink&lsqb;30&rsqb;_FO )_) ._. 
</s>
<s>
We_PPIS2 have_VH0 proposed_VVN a_AT1 formal_JJ specification_NN1 of_IO all_DB metadata_NN1 that_CST enters_VVZ the_AT network_NN1 ._. 
</s>
<s>
This_DD1 allows_VVZ us_PPIO2 to_TO treat_VVI OpenADN_NN1 messages_NN2 to_TO be_VBI one_MC1 datagram_NN1 long_RR ._. 
</s>
<s>
This_DD1 focus_NN1 on_II edges_NN2 is_VBZ supported_VVN by_II the_AT success_NN1 of_IO previous_JJ risk_NN1 models_NN2 applying_VVG a_AT1 similar_JJ approach_NN1 (_( Blennow_NP1 and_CC Salln_VV0 ?_? s_ZZ1 ,_, 2004_MC ,_, Blennow_NP1 et_NP1 ?_? al._NNU ,_, 2010_MC )_) ._. 
</s>
<s>
Each_DD1 block_NN1 does_VDZ not_XX overlap_VVI with_IW the_AT others_NN2 ._. 
</s>
<s>
Finally_RR ,_, set_NN1 τ_NULL (_( ai_NNU )_) =i_FO for_IF all_DB i>2_FO 2t_FO ._. 
</s>
<s>
Part_NN1 correspondence_NN1 can_VM be_VBI viewed_VVN as_II a_AT1 type_NN1 of_IO shape_NN1 comparison_NN1 intermediate_JJ between_II geometric_JJ and_CC structural_JJ comparisons_NN2 ._. 
</s>
<s>
Higher_JJR flexibility_NN1 in_II the_AT pricing_NN1 model_NN1 will_VM increase_VVI the_AT revenue_NN1 generation_NN1 by_II the_AT Web_NN1 services_NN2 provider_NN1 ._. 
</s>
<s>
This_DD1 is_VBZ despite_II regular_JJ updates_NN2 to_II the_AT filter_NN1 rule_NN1 set_NN1 and_CC software_NN1 ._. 
</s>
<s>
Note_VV0 that_CST we_PPIS2 performed_VVD similar_JJ performance_NN1 measurement_NN1 and_CC evaluation_NN1 steps_NN2 on_II the_AT other_JJ four_MC RT_JJ constraints_NN2 in_II SCAPS_NN2 (_( i.e._REX ,_, HRTC1_FO ,_, HRTC2_FO ,_, HRTC3_FO ,_, andSRTC2_FO )_) which_DDQ due_JJ to_TO space_VVI constraints_NN2 are_VBR not_XX reported_VVN in_II this_DD1 article_NN1 ._. 
</s>
<s>
The_AT notable_JJ thing_NN1 about_II the_AT survey_NN1 is_VBZ the_AT table_NN1 with_IW title_NN1 '_NULL The_AT cost_NN1 of_IO Computer_NN1 Crime_NN1 '_NULL ,_, where_CS one_PN1 can_VM view_VVI aggregate_JJ costs_NN2 sampled_VVN over_II a_AT1 48_MC month_NNT1 period_NN1 (_( 2000–2003_MCMC )_) ._. 
</s>
<s>
Fig._NN1 5_MC shows_VVZ linearly_RR increasing_JJ ingestion_NN1 rate_NN1 for_IF each_DD1 consecutive_JJ version_NN1 for_IF BEAR-A_NN1 ,_, while_CS Fig._NN1 6_MC shows_NN2 corresponding_VVG linearly_RR increasing_JJ storage_NN1 sizes_NN2 ._. 
</s>
<s>
APPELis_NN1 a_AT1 family_NN1 of_IO policy_NN1 languages_NN2 with_IW a_AT1 common_JJ core_NN1 ._. 
</s>
<s>
If_CS the_AT trend_NN1 we_PPIS2 have_VH0 seen_VVN continues_VVZ ,_, then_RT virtually_RR all_DB trees_NN2 of_IO a_AT1 small_JJ size_NN1 will_VM have_VHI a_AT1 unique_JJ Laplacian_JJ spectrum_NN1 ._. 
</s>
<s>
The_AT Semantic_JJ Web_NN1 community_NN1 has_VHZ been_VBN creating_VVG large-scale_JJ knowledge_NN1 graphs_NN2 defining_VVG a_AT1 multitude_NN1 of_IO entity_NN1 types_NN2 ._. 
</s>
<s>
Conceptual_JJ level_JJ standards_NN2 to_TO support_VVI business_NN1 analysts_NN2 in_II the_AT expression_NN1 and_CC implementation_NN1 of_IO signature_NN1 functionality_NN1 within_II proactive_JJ fraud_NN1 policy_NN1 controls_NN2 are_VBR therefore_RR vital_JJ advancements_NN2 towards_II supporting_VVG the_AT effective_JJ deployment_NN1 of_IO signature_NN1 functionality_NN1 and_CC increasing_VVG an_AT1 organisations_NN2 responsiveness_NN1 to_II rapidly_RR evolving_VVG fraudster_NN1 behaviour_NN1 ._. 
</s>
<s>
A_AT1 number_NN1 of_IO methods_NN2 are_VBR proposed_VVN and_CC validated_VVN for_IF aligning_VVG images_NN2 of_IO the_AT same_DA modality_NN1 ,_, also_RR known_VVN as_II intramodal_JJ registration_NN1 ._. 
</s>
<s>
Our_APPGE focus_NN1 in_II this_DD1 paper_NN1 is_VBZ the_AT continuous_JJ implicit_JJ authentication_NN1 protocol_NN1 between_II the_AT carrier_NN1 and_CC the_AT device_NN1 ._. 
</s>
<s>
It_PPH1 is_VBZ our_APPGE belief_NN1 that_CST each_DD1 of_IO the_AT four-phases_NN2 of_IO KM-competence_NN1 must_VM be_VBI addressed_VVN in_II all_DB lifecycle-wide_JJ management_NN1 plans_NN2 for_IF Enterprise_NN1 Systems_NN2 ._. 
</s>
<s>
In_II this_DD1 section_NN1 ,_, we_PPIS2 develop_VV0 a_AT1 framework_NN1 that_CST assesses_VVZ the_AT suitability_NN1 of_IO a_AT1 MAS_JJ solution_NN1 given_VVN a_AT1 specific_JJ problem_NN1 ._. 
</s>
<s>
Features_NN2 derived_VVN from_II packet_NN1 contents_NN2 (_( rather_II21 than_II22 headers_NN2 )_) are_VBR required_VVN to_TO reliably_RR detect_VVI these_DD2 attacks_NN2 ._. 
</s>
<s>
Modelling_NN1 and_CC water_NN1 quality_NN1 monitoring_NN1 research_NN1 and_CC application_NN1 of_IO the_AT SedNet_NN1 and_CC ANNEX_VV0 models_NN2 in_II the_AT region_NN1 were_VBDR undertaken_VVN to_TO evaluate_VVI sediment_NN1 and_CC nutrient_NN1 sources_NN2 and_CC loads_NN2 from_II 33_MC freshwater_JJ catchment_NN1 management_NN1 areas_NN2 ,_, to_TO support_VVI development_NN1 of_IO WQIP_NP1 water_NN1 quality_NN1 objectives_NN2 and_CC targets_NN2 ,_, and_CC evaluation_NN1 of_IO likely_JJ effects_NN2 of_IO water_NN1 quality_NN1 on_II improved_JJ and_CC improved_JJ management_NN1 practices_VVZ scenarios_NN2 and_CC improved_JJ management_NN1 practice_NN1 adoption_NN1 rates_NN2 ._. 
</s>
<s>
It_PPH1 is_VBZ likely_JJ that_CST 500_MC topics_NN2 are_VBR too_RG many_DA2 for_IF this_DD1 version_NN1 of_IO Mozilla_NP1 ._. 
</s>
<s>
After_CS several_DA2 iterations_NN2 in_II this_DD1 phase_NN1 of_IO the_AT CPR_NP1 process_NN1 ,_, a_AT1 suitable_JJ set_NN1 of_IO strategies_NN2 were_VBDR identified_VVN and_CC implementation_NN1 plans_NN2 were_VBDR made_VVN ._. 
</s>
<s>
This_DD1 model_NN1 includes_VVZ the_AT golden_JJ triangle_NN1 ,_, but_CCB includes_VVZ three_MC other_JJ components_NN2 to_TO assess_VVI :_: the_AT information_NN1 system_NN1 developed_VVD ,_, including_II its_APPGE reliability_NN1 and_CC maintainability_NN1 ;_; benefits_NN2 to_II the_AT organization_NN1 ,_, such_II21 as_II22 improved_JJ efficiency_NN1 and_CC effectiveness_NN1 ,_, learning_VVG and_CC profits_NN2 ;_; and_CC benefits_NN2 to_II stakeholders_NN2 ,_, such_II21 as_II22 user_NN1 satisfaction_NN1 ,_, environmental_JJ impact_NN1 ,_, personal_JJ development_NN1 and_CC professional_JJ learning_NN1 &lsqb;_( 6_MC &rsqb;_) ._. 
</s>
<s>
This_DD1 research_NN1 achieved_VVN convincing_JJ results_NN2 featuring_VVG efficient_JJ covering_NN1 of_IO the_AT network_NN1 by_II a_AT1 minimal_JJ number_NN1 of_IO sensors_NN2 ._. 
</s>
<s>
All_DB of_IO the_AT ordinary_JJ A*_FO algorithms_NN2 use_VV0 the_AT same_DA strategy_NN1 as_CSA GHSETA*_FO to_TO eliminate_VVI duplicates_NN2 ._. 
</s>
<s>
The_AT aim_NN1 of_IO this_DD1 part_NN1 of_IO our_APPGE study_NN1 is_VBZ to_TO simply_RR explore_VVI the_AT effect_NN1 of_IO these_DD2 methods_NN2 on_II the_AT prediction_NN1 accuracy_NN1 ,_, and_CC not_XX to_TO prove_VVI the_AT superiority_NN1 of_IO one_PN1 over_II the_AT other_JJ ._. 
</s>
<s>
Flux_NN1 chambers_NN2 provide_VV0 only_RR a_AT1 snapshot_NN1 of_IO emissions_NN2 ,_, not_XX only_RR in_II space_NN1 ,_, but_CCB also_RR in_II time_NNT1 ._. 
</s>
<s>
AI_NN1 programs_NN2 capable_JJ of_IO perfect_JJ play_NN1 exist_VV0 ,_, such_II21 as_II22 Chinook&lsqb;70&rsqb;_FO ._. 
</s>
<s>
First_MD ,_, we_PPIS2 import_VV0 a_AT1 shapefile_NN1 containing_VVG USA_NP1 second-level_JJ administrative_JJ divisions_NN2 to_II a_AT1 PostGIS_NN1 database_NN1 ._. 
</s>
<s>
While_CS the_AT exercise_NN1 regimen_NN1 is_VBZ able_JK to_TO improve_VVI Sarah_NP1 '_NULL s_ZZ1 physical_JJ health_NN1 ,_, it_PPH1 should_VM be_VBI noted_VVN that_CST other_JJ interventions_NN2 can_VM yield_VVI even_RR greater_JJR benefits_NN2 ._. 
</s>
<s>
For_IF all_DB three_MC synthetic_JJ models_NN2 ,_, the_AT error_NN1 term_NN1 ?_? 
</s>
<s>
Then_RT ,_, a_AT1 Greedy_JJ Filtering_NN1 with_IW threshold_NN1 θ_NULL =0.2716023_FO is_VBZ applied_VVN on_II these_DD2 mappings_NN2 to_TO obtain_VVI the_AT final_JJ result_NN1 as_CSA shown_VVN in_II Table_NN1 5_MC ._. 
</s>
<s>
This_DD1 has_VHZ been_VBN addressed_VVN earlier_JJR but_CCB only_RR for_IF Group_NN1 IV_MC &lsqb;_( 74_MC &rsqb;_) ._. 
</s>
<s>
However_RR ,_, it_PPH1 took_VVD more_DAR than_CSN ten_MC months_NNT2 to_TO fix_VVI the_AT bug_NN1 ._. 
</s>
<s>
The_AT contribution_NN1 of_IO this_DD1 paper_NN1 is_VBZ twofold_RR :_: firstly_RR the_AT paper_NN1 presents_VVZ the_AT first_MD thorough_JJ literature_NN1 survey_NN1 on_II the_AT topic_NN1 of_IO tagging_VVG motivation_NN1 ._. 
</s>
<s>
The_AT eight_MC views_NN2 of_IO the_AT toy_NN1 duck_NN in_II Fig._NN1 11_MC are_VBR taken_VVN consecutively_RR with_IW interval_NN1 5°_NNU ._. 
</s>
<s>
The_AT first_MD option_NN1 is_VBZ to_TO set_VVI it_PPH1 to_II IPA_NP1 ._. 
</s>
<s>
Thus_RR even_RR in_II cases_NN2 where_RRQ Method_NN1 2_MC required_VVN fewer_DAR subdomain_NN1 solves_VVZ ,_, Method_NN1 3_MC was_VBDZ more_RGR efficient_JJ in_II31 terms_II32 of_II33 CPU_NN1 time_NNT1 ,_, as_CSA the_AT time_NNT1 for_IF the_AT coarse_JJ solves_VVZ was_VBDZ not_XX negligible_JJ ._. 
</s>
<s>
In_II &lsqb;_( 10_MC &rsqb;_) ,_, we_PPIS2 have_VH0 described_VVN a_AT1 practical_JJ way_NN1 to_TO deploy_VVI DRES_NN2 in_II an_AT1 incremental_JJ manner_NN1 such_CS21 that_CS22 there_EX are_VBR significant_JJ benefits_NN2 over_II conventional_JJ reservations_NN2 ._. 
</s>
<s>
Once_CS there_EX is_VBZ an_AT1 active_JJ node_NN1 in_II W2_FO ,_, W1_FO does_VDZ not_XX present_VVI a_AT1 pure_JJ list_NN1 of_IO concepts_NN2 any_DD more_DAR ,_, but_CCB it_PPH1 lists_VVZ object_NN1 property_NN1 and_CC range_NN1 concept_NN1 pairs_NN2 pertaining_II21 to_II22 the_AT active_JJ node_NN1 ._. 
</s>
<s>
That_REX21 is_REX22 ,_, given_VVN any_DD non-convex_JJ aggregate_JJ context_NN1 ,_, we_PPIS2 can_VM use_VVI it_PPH1 to_TO simulate_VVI implications_NN2 ,_, thus_RR to_TO simulate_VVI disjunctive_JJ rules_NN2 ._. 
</s>
<s>
Users_NN2 should_VM be_VBI familiar_JJ with_IW these_DD2 issues_NN2 and_CC possible_JJ limitations_NN2 in_II using_VVG downscaled_JJ outputs_NN2 (_( see_VV0 Timbal_NP1 ,_, 2006a_FO ,_, for_IF the_AT particular_JJ case_NN1 of_IO the_AT SDM_NP1 presented_VVD here_RL )_) as_II31 well_II32 as_II33 being_VBG aware_JJ of_IO the_AT important_JJ benefits_NN2 brought_VVN by_II SDMs_NP1 compared_VVD with_IW using_VVG Direct_JJ Model_NN1 Outputs_NN2 (_( DMOs_NP2 )_) ._. 
</s>
<s>
Providing_VVG multiple_JJ routes_NN2 for_IF accessing_VVG Revyu_NN1 data_NN (_( Javascript_NP1 ,_, RSS_NP1 ,_, RDF_NP1 ,_, and_CC SPARQL_NP1 )_) allows_VVZ site_NN1 users_NN2 to_II easily_RR syndicate_NN1 reviews_VVZ from_II the_AT site_NN1 for_IF reuse_NN1 in_II their_APPGE own_DA applications_NN2 ._. 
</s>
<s>
Savolainen_NP1 et_RA21 al_RA22 ._. 
</s>
<s>
(_( Savolainen_NP1 et_RA21 al._RA22 ,_, 2007_MC )_) focus_VV0 on_II information_NN1 security_NN1 for_IF service_NN1 centric_JJ systems_NN2 ._. 
</s>
<s>
In_BCL21 order_BCL22 to_TO use_VVI unintended_JJ USB_NN1 channels_NN2 to_TO exfiltrate_VVI data_NN from_II a_AT1 network_NN1 endpoint_NN1 ,_, a_AT1 Hardware_NN1 Trojan_JJ would_VM have_VHI to_TO be_VBI successfully_RR enumerated_VVN by_II a_AT1 network_NN1 endpoint_NN1 ._. 
</s>
<s>
Section_NN1 3.2_MC ,_, allowing_VVG more_RGR efficient_JJ search_NN1 ._. 
</s>
<s>
For_IF two_MC systems_NN2 with_IW the_AT same_DA mission_NN1 and_CC purpose_NN1 ,_, the_AT performance_NN1 ,_, the_AT vulntest_JJT and_CC the_AT resilience_NN1 requirements_NN2 may_VM be_VBI expected_VVN to_TO be_VBI similar_JJ enough_RR such_CS21 that_CS22 the_AT best_JJT metric_JJ score_NN1 in_II each_DD1 of_IO these_DD2 three_MC areas_NN2 would_VM become_VVI the_AT 100%_NNU mark_NN1 for_IF the_AT purposed_JJ of_IO STAC_NN1 ._. 
</s>
<s>
In_II the_AT first_MD numerical_JJ experiment_NN1 ,_, the_AT one-dimensional_JJ Fokker–Planck_NN1 equation_NN1 is_VBZ solved_VVN in_II the_AT space–time_NNT1 domain_NN1 with_IW T=10T=10_FO ._. 
</s>
<s>
Our_APPGE research_NN1 uses_VVZ discourse_NN1 analytic_JJ techniques_NN2 to_TO examine_VVI the_AT ways_NN2 that_CST CISOs_NN2 communicate_VV0 information_NN1 security_NN1 requirements_NN2 ._. 
</s>
<s>
The_AT decision_NN1 to_TO adopt_VVI this_DD1 approach_NN1 was_VBDZ made_VVN in_II31 light_II32 of_II33 Requirement_NN1 AQ1_FO to_TO reuse_VVI existing_JJ capabilities_NN2 ,_, in_II this_DD1 case_NN1 ,_, the_AT already_RR existing_JJ HTML_NP1 link_NN1 element_NN1 ._. 
</s>
<s>
In_RR21 addition_RR22 ,_, as_CSA the_AT market_NN1 for_IF component-based_JJ software_NN1 heats_NN2 up_RP ,_, many_DA2 standard_JJ business_NN1 application_NN1 logic_NN1 components_NN2 will_VM be_VBI available_JJ for_IF purchase_NN1 off_II the_AT shelf_NN1 ._. 
</s>
<s>
Strategic_JJ IT_NN1 system_NN1 development_NN1 outsourcing_NN1 is_VBZ now_RT being_VBG pursued_VVN actively_RR by_II vendors_NN2 and_CC ,_, because_II21 of_II22 the_AT importance_NN1 of_IO such_DA projects_NN2 to_II the_AT client_NN1 organization_NN1 ,_, unaddressed_JJ risks_NN2 that_CST develop_VV0 into_II problems_NN2 can_VM have_VHI seriously_RR detrimental_JJ effects_NN2 on_II the_AT organization_NN1 ._. 
</s>
<s>
Litterfall_VV0 in_II coniferous_JJ forests_NN2 is_VBZ assumed_VVN to_TO be_VBI constant_JJ for_IF each_DD1 day_NNT1 of_IO the_AT year_NNT1 (_( Pedersen_NP1 and_CC Bille-Hansen_NP1 ,_, 1999_MC )_) ,_, with_IW the_AT daily_JJ rate_NN1 reset_VV0 each_DD1 year_NNT1 depending_VVG the_AT previous_JJ year_NNT1 '_NULL s_ZZ1 maximum_JJ LAI_NP1 and_CC a_AT1 species-specific_JJ turnover_NN1 rate_NN1 ._. 
</s>
<s>
We_PPIS2 first_MD prove_VV0 that_CST P_ZZ1 contains_VVZ the_AT newly_RR inferred_VVN permissions_NN2 as_CSA per_II Definition_NN1 3_MC ,_, and_CC then_RT we_PPIS2 show_VV0 an_AT1 example_NN1 of_IO how_RRQ the_AT Inference_NN1 CP-net_NN1 works_VVZ ._. 
</s>
<s>
The_AT ICARUS_NP1 framework_NN1 introduced_VVN in_II this_DD1 section_NN1 allows_VVZ us_PPIO2 to_TO define_VVI and_CC analyse_VVI such_DA enhancements_NN2 and_CC their_APPGE combinations_NN2 in_II an_AT1 instructive_JJ ,_, formal_JJ and_CC consistent_JJ way_NN1 ._. 
</s>
<s>
This_DD1 computation_NN1 is_VBZ ,_, in_II fact_NN1 ,_, a_AT1 reduction_NN1 problem_NN1 ,_, and_CC can_VM not_XX be_VBI solved_VVN using_VVG just_RR one_MC1 kernel_NN1 invocation_NN1 (_( because_CS the_AT thread_NN1 blocks_NN2 can_VM not_XX synchronize_VVI their_APPGE execution_NN1 )_) ._. 
</s>
<s>
For_IF a_AT1 full_JJ definition_NN1 of_IO OWL_NN1 2_MC ,_, please_RR refer_VV0 to_II the_AT OWL_NN1 2_MC Structural_JJ Specification_NN1 and_CC Direct_JJ Semantics_NN1 &lsqb;_( 3_MC &rsqb;_) ,_, &lsqb;_( 4_MC &rsqb;_) ;_; here_RL we_PPIS2 just_RR recapitulate_VV0 the_AT relevant_JJ terminology_NN1 ._. 
</s>
<s>
One_MC1 of_IO the_AT dicta_NN2 of_IO consumer_NN1 health_NN1 information_NN1 is_VBZ the_AT model_NN1 of_IO an_AT1 informed_JJ decision_NN1 maker_NN1 weighing_VVG conflicting_JJ choices_NN2 in_II31 consultation_II32 with_II33 a_AT1 doctor_NN1 (_( Spink_NP1 et_RA21 al._RA22 ,_, 2004_MC )_) ;_; but_CCB while_CS some_DD patients_NN2 prefer_VV0 this_DD1 model_NN1 ,_, many_DA2 would_VM just_RR as_RG soon_RR not_XX be_VBI presented_VVN with_IW choices_NN2 or_CC conflicting_JJ information_NN1 (_( Rimal_NP1 ,_, 2001_MC )_) and_CC the_AT pragmatic_JJ limits_NN2 of_IO an_AT1 eight_MC minute_JJ consultation_NN1 in_II the_AT U.S._NP1 health_NN1 care_NN1 systems_NN2 make_VV0 it_PPH1 impractible_JJ in_II any_DD case_NN1 (_( Henwood_NP1 ,_, Wyatt_NP1 ,_, Hart_NP1 ,_, &;_NULL Smith_NP1 ,_, 2003_MC )_) ._. 
</s>
<s>
Evaluation_NN1 1_MC1 :_: a_AT1 task_NN1 of_IO finding_VVG videos_NN2 of_IO political_JJ figures_NN2 of_IO 2008_MC ._. 
</s>
<s>
In_II Definition_NN1 2_MC ,_, the_AT set_NN1 of_IO inference_NN1 tuples_NN2 ,_, I_PPIS1 ,_, that_CST are_VBR applicable_JJ to_II an_AT1 organization_NN1 were_VBDR defined_VVN ._. 
</s>
<s>
The_AT net_JJ present_JJ value_NN1 is_VBZ expressed_VVN as_II a_AT1 percentage_NN1 of_IO GDP_NN1 in_II year_NNT1 1_MC1 ._. 
</s>
<s>
While_CS CSP_NP1 and_CC SAT-based_JJ approaches_NN2 ,_, including_II CSP-based_JJ planning_NN1 &lsqb;_( 16_MC &rsqb;_) ,_, are_VBR typically_RR based_VVN on_II the_AT discretisation_NN1 of_IO time_NNT1 ,_, AI_NN1 planning_NN1 can_VM exploit_VVI continuous_JJ time_NNT1 ._. 
</s>
<s>
As_II a_AT1 special_JJ case_NN1 ,_, more_RGR complex_JJ policies_NN2 may_VM also_RR be_VBI defined_VVN by_II network_NN1 operators_NN2 ._. 
</s>
<s>
We_PPIS2 distinguish_VV0 in_II our_APPGE analysis_NN1 between_II different_JJ types_NN2 of_IO verification_NN1 problems_NN2 (_( or_CC questions_NN2 )_) ,_, which_DDQ we_PPIS2 selected_VVD based_VVN on_II our_APPGE extensive_JJ experience_NN1 with_IW teaching_VVG OWL_NN1 and_CC ontology_NN1 authoring_VVG &lsqb;_( 9_MC &rsqb;_) ._. 
</s>
<s>
For_IF each_DD1 triangle_NN1 ,_, two_MC of_IO its_APPGE vertices_NN2 (_( CDT_NP1 points_VVZ )_) will_VM lie_VVI on_II one_MC1 contour_NN1 ,_, while_CS the_AT third_MD lies_NN2 on_II the_AT opposite_JJ contour_NN1 ._. 
</s>
<s>
Missier_NN1 and_CC coworkers_NN2 propose_VV0 with_IW Janus_NP1 a_AT1 domain-aware_JJ provenance_NN1 model_NN1 by_II extending_VVG the_AT Provenir_NN1 upper-level_NN1 ontology_NN1 &lsqb;_( 33_MC &rsqb;_) grounded_VVD to_II BFO_NP1 &lsqb;_( 34_MC &rsqb;_) (_( Basic_JJ Formal_JJ Ontology_NN1 )_) concepts_NN2 ,_, and_CC a_AT1 prototype_NN1 implementation_NN1 within_II the_AT Taverna_NP1 workflow_NN1 workbench_NN1 ._. 
</s>
<s>
In_II future_JJ calls_NN2 to_II the_AT propagator_NN1 ,_, the_AT two_MC SCCs_NP1 can_VM be_VBI considered_VVN independently_RR ._. 
</s>
<s>
Special_JJ emphasis_NN1 is_VBZ given_VVN to_II the_AT following_JJ six_MC main_JJ components_NN2 of_IO any_DD measurement_NN1 system_NN1 :_: (_( i_ZZ1 )_) Accuracy_NN1 (_( ii_MC )_) robustness_NN1 (_( iii_MC )_) validity_NN1 (_( iv_MC )_) functionality_NN1 (_( v_ZZ1 )_) time_VV0 and_CC (_( vi_MC )_) costs_VVZ ._. 
</s>
<s>
Following_VVG the_AT scale_NN1 refinement_NN1 phase_NN1 ,_, we_PPIS2 evaluated_VVD the_AT internal_JJ consistency_NN1 of_IO each_DD1 first-order_JJ factor_NN1 by_II examining_VVG the_AT estimates_NN2 of_IO composite_JJ reliability_NN1 and_CC variance_NN1 (_( Hair_NN1 et_RA21 al._RA22 ,_, 1998_MC )_) ._. 
</s>
<s>
Rule_NN1 7_MC ._. 
</s>
<s>
When_CS aT-node_JJ changes_NN2 status_NN1 toinsystem_NN1 ,_, it_PPH1 must_VM inform_VVI all_DB its_APPGE reverse-neighbors_NN2 (_( by_II sendingInSysNotiMsg_NN1 )_) ,_, in_II31 addition_II32 to_II33 its_APPGE neighbors_NN2 ,_, that_CST it_PPH1 has_VHZ become_VVN anS-node_NN1 ._. 
</s>
<s>
Particularly_RR for_IF words_NN2 like_II "_" trust_NN1 "_" or_CC "_" risk_NN1 ,_, "_" two_MC different_JJ disciplines_NN2 can_VM use_VVI the_AT same_DA word_NN1 but_CCB with_IW very_RG different_JJ meanings_NN2 and_CC assumptions_NN2 ._. 
</s>
<s>
Resources_NN2 in_II the_AT "_" ns_FO "_" class_NN1 are_VBR allowed_VVN to_TO use_VVI fragment_NN1 identifiers_NN2 ._. 
</s>
<s>
Actors_NN2 also_RR define_VV0 one_PPX121 another_PPX122 in_II their_APPGE interaction_NN1 (_( Callon_NP1 ,_, 1991_MC )_) ._. 
</s>
<s>
Before_II conducting_VVG an_AT1 objective_JJ evaluation_NN1 ,_, a_AT1 brief_JJ description_NN1 of_IO a_AT1 few_DA2 binarization_NN1 methods_NN2 used_VVN for_IF this_DD1 purpose_NN1 is_VBZ provided_VVN ._. 
</s>
<s>
Instead_RR ,_, one_MC1 of_IO the_AT ten_MC Qualifications_NN2 should_VM be_VBI used_VVN ;_; in_II that_DD1 sense_NN1 ,_, the_AT influence_NN1 relation_NN1 is_VBZ "_" abstract_JJ "_" ._. 
</s>
<s>
The_AT combination_NN1 of_IO two_MC model_NN1 components_NN2 (_( GUIDOS/MSPA_FU application_NN1 and_CC Landscape_VV0 Mosaic_JJ model_NN1 )_) is_VBZ a_AT1 new_JJ tool_NN1 to_TO provide_VVI the_AT edge_NN1 interface_NN1 context_NN1 of_IO identified_JJ morphological_JJ shapes_NN2 (_( edge_NN1 of_IO interior_JJ habitat_NN1 ,_, connecting_VVG linear_JJ features_NN2 and_CC physically_RR isolated_JJ islets_NN2 )_) ._. 
</s>
<s>
The_AT device_NN1 confidence_NN1 ,_, represented_VVN over_II time_NNT1 by_II the_AT solid_JJ line_NN1 ,_, first_MD begins_VVZ at_II 0.5_MC at_II the_AT far_JJ left_NN1 of_IO the_AT diagram_NN1 ._. 
</s>
<s>
Automated_JJ approaches_NN2 are_VBR simple_JJ to_TO use_VVI (_( especially_RR Guan_JJ '_NULL s_ZZ1 )_) ,_, but_CCB the_AT resulting_JJ test_NN1 goals_NN2 are_VBR currently_RR formulated_VVN in_II a_AT1 format_NN1 such_II21 as_II22 LOTOStraces_NN2 less_RGR flexible_JJ than_CSN XML_NP1 ._. 
</s>
<s>
The_AT model_NN1 shows_VVZ that_CST when_CS the_AT negative_JJ influence_NN1 (_( β_NULL )_) is_VBZ fixed_VVN ,_, increasing_VVG positive_JJ influence_NN1 (_( )_) to_II a_AT1 certain_JJ threshold_NN1 eliminates_VVZ the_AT HRP_NP1 population_NN1 ._. 
</s>
<s>
Fig._NN1 13(f)_FO gives_VVZ the_AT segmentation_NN1 results_NN2 of_IO the_AT proposed_JJ scheme_NN1 ,_, which_DDQ also_RR detects_VVZ the_AT comparative_JJ boundaries_NN2 of_IO the_AT two_MC mass_JJ regions_NN2 ._. 
</s>
<s>
We_PPIS2 then_RT increase_VV0 w_ZZ1 by_II 5%_NNU and_CC 10%_NNU of_IO the_AT length_NN1 of_IO the_AT longest_JJT sequence_NN1 to_TO observe_VVI the_AT affect_NN1 of_IO larger_JJR windows_NN2 sizes_NN2 on_II the_AT approaches_NN2 ._. 
</s>
<s>
Applying_VVG GR2_FO again_RT tells_VVZ us_PPIO2 that_CST ai+1_FO is_VBZ a_AT1 customer_NN1 of_IO ai_NNU for_IF each_DD1 i=p+1t-1_FO ._. 
</s>
<s>
To_TO select_VVI articles_NN2 that_CST address_VV0 a_AT1 specific_JJ research_NN1 question_NN1 ,_, P50_FO ,_, P56_FO ._. 
</s>
<s>
If_CS this_DD1 would_VM be_VBI the_AT case_NN1 ,_, unproductive_JJ firms_NN2 are_VBR not_XX bankrupted_VVN ,_, and_CC can_VM return_VVI to_II producing_VVG goods_NN2 in_II the_AT subsequent_JJ periods_NN2 ._. 
</s>
<s>
The_AT dislocation_NN1 dynamics_NN code_VV0 again_RT writes_VVZ dislocation_NN1 segment_NN1 information_NN1 to_II DSM_NP1 ,_, consisting_VVG now_RT of_IO eight_MC doubles_NN2 per_II segment_NN1 :_: vertex_VV0 coordinates_NN2 and_CC a_AT1 flag_NN1 for_IF each_DD1 vertex_NN1 indicating_VVG whether_CSW it_PPH1 has_VHZ been_VBN previously_RR marked_VVN as_II a_AT1 vertex_NN1 in_II B._NP1 Then_RT ,_, the_AT dislocation_NN1 dynamics_NN code_NN1 releases_VVZ control_NN1 of_IO DSM_NP1 and_CC the_AT finite_JJ element_NN1 code_NN1 reads_VVZ the_AT dislocation_NN1 line_NN1 segment_NN1 data_NN and_CC begins_VVZ updating_VVG dislocation_NN1 line_NN1 segment_NN1 information_NN1 to_TO account_VVI for_IF bounded_JJ B._NP1 We_PPIS2 validated_VVD the_AT clustering_NN1 performance_NN1 of_IO all_DB competing_JJ methods_NN2 using_VVG real_JJ data_NN of_IO gene_NN1 expressions_NN2 and_CC metabolic_JJ pathways_NN2 by_II running_VVG each_DD1 method_NN1 100_MC times_NNT2 ._. 
</s>
<s>
In_II the_AT context_NN1 of_IO repartitioning_VVG this_DD1 can_VM be_VBI rewritten_VVN as_II where_RRQ denotes_VVZ the_AT new_JJ sub-domain_NN1 assigned_VVN to_TO process_VVI p_ZZ1 ._. 
</s>
<s>
Finally_RR ,_, in_BCL21 order_BCL22 to_TO encompass_VVI different_JJ spatial_JJ frequencies_NN2 ,_, localities_NN2 and_CC orientation_NN1 selectivities_NN2 ,_, we_PPIS2 combined_VVD all_DB these_DD2 features_NN2 to_II a_AT1 single_JJ feature_NN1 vector_NN1 by_II concatenating_VVG all_DB these_DD2 representations_NN2 ,_, defined_VVN as_CSA (_( (_( O0,0()_FO )_) T_ZZ1 (_( O0,1_NP1 (_( )_) )_) T_ZZ1 (_( O4,7_NP1 (_( )_) )_) T_ZZ1 )_) T_ZZ1 ,_, where_CS T_ZZ1 was_VBDZ a_AT1 transpose_VV0 operator_NN1 ._. 
</s>
<s>
The_AT horizontal_JJ and_CC vertical_JJ structure_NN1 of_IO the_AT solution_NN1 frame_NN1 is_VBZ shown_VVN schematically_RR in_II Table_NN1 1_MC1 ,_, where_CS for_IF the_AT first_MD column_NN1 each_DD1 element_NN1 in_II a_AT1 table_NN1 cell_NN1 corresponds_VVZ to_II an_AT1 instance_NN1 of_IO a_AT1 meta-class_JJ from_II the_AT meta-model_NN1 in_II Fig._NN1 2_MC (_( we_PPIS2 omit_VV0 showing_VVG this_DD1 explicitly_RR on_II the_AT table_NN1 for_IF the_AT sake_NN1 of_IO readability_NN1 )_) ._. 
</s>
<s>
The_AT implementation_NN1 and_CC feasibility_NN1 of_IO using_VVG a_AT1 MD_NN1 framework_NN1 for_IF solving_VVG an_AT1 inverse_JJ problem_NN1 has_VHZ been_VBN demonstrated_VVN both_RR on_II conventional_JJ and_CC GPGPU_NP1 hardware_NN1 and_CC for_IF a_AT1 multi-core_JJ CPU_NN1 for_IF the_AT specific_JJ problem_NN1 of_IO a_AT1 2D_NNU fracture_NN1 problem_NN1 on_II the_AT atomic_JJ level_NN1 ._. 
</s>
<s>
However_RR ,_, a_AT1 field_NN1 that_CST includes_VVZ the_AT word_NN1 "_" engineering_NN1 "_" can_VM not_XX be_VBI purely_RR theoretical_JJ ._. 
</s>
<s>
In_BCL21 order_BCL22 for_IF a_AT1 semantic_JJ application_NN1 to_TO consume_VVI this_DD1 data_NN ,_, it_PPH1 has_VHZ to_TO know_VVI what_DDQ the_AT data_NN means_NN ;_; it_PPH1 is_VBZ not_XX useful_JJ knowing_VVG that_CST you_PPY can_VM search_VVI for_IF books_NN2 via_II Amazon.com_NP1 '_NULL s_ZZ1 E-Commerce_NP1 Web_NN1 Service_NN1 if_CS the_AT application_NN1 consuming_VVG this_DD1 data_NN doesn_VV0 '_NULL t_ZZ1 know_VV0 that_CST the_AT service_NN1 returns_VVZ a_AT1 book_NN1 or_CC even_RR how_RRQ a_AT1 book_NN1 is_VBZ defined_VVN ._. 
</s>
<s>
The_AT first_MD approach_NN1 is_VBZ to_TO expose_VVI the_AT structured_JJ data_NN that_CST already_RR underlies_VVZ the_AT unstructured_JJ web_NN1 pages_NN2 ._. 
</s>
<s>
The_AT only_JJ robot–robot_NN1 collision_NN1 occurs_VVZ at_II (_( B1_FO ,_, B1_FO ,_, B3_FO )_) ,_, and_CC the_AT involved_JJ robots_NN2 have_VH0 already_RR been_VBN added_VVN to_II the_AT collision_NN1 set_NN1 of_IO the_AT initial_JJ configuration_NN1 ,_, the_AT only_JJ state_NN1 in_II the_AT backpropagation_NN1 set_NN1 of_IO (_( B1_FO ,_, B1_FO ,_, B3_FO )_) ._. 
</s>
<s>
The_AT receiver_NN1 operating_NN1 characteristics_NN2 (_( ROC_NP1 )_) of_IO these_DD2 methods_NN2 are_VBR presented_VVN in_II Figs._NN2 4(a)_FO and_CC (_( b_ZZ1 )_) ._. 
</s>
<s>
This_DD1 issue_NN1 is_VBZ explored_VVN in_II Section_NN1 8_MC ._. 
</s>
<s>
Skeletonisation_NN1 artifacts_NN2 ,_, described_VVN in_II more_DAR detail_NN1 in_II Ref_NN1 ._. 
</s>
<s>
However_RR ,_, because_CS such_DA an_AT1 analysis_NN1 is_VBZ space-intensive_JJ ,_, we_PPIS2 present_VV0 a_AT1 full_JJ discussion_NN1 of_IO only_RR the_AT three_MC selected_JJ bugs_NN2 ,_, which_DDQ are_VBR representative_JJ of_IO different_JJ situations_NN2 that_CST can_VM result_VVI from_II using_VVG our_APPGE approach_NN1 ._. 
</s>
<s>
In_RR21 general_RR22 ,_, both_DB2 ?_? rMOCF_NN1 and_CC ?_? gMOCF_NN1 are_VBR larger_JJR for_IF ?_? 
</s>
<s>
SF_NN1 ?_? networks_NN2 than_CSN for_IF ?_? 
</s>
<s>
ED_NP1 ?_? networks_VVZ ,_, for_IF a_AT1 given_JJ ?_? ?_? ._. 
</s>
<s>
Our_APPGE BBN_NP1 architecture_NN1 provides_VVZ a_AT1 framework_NN1 for_IF processing_VVG multiple_JJ biometrics_NN2 and_CC per-score_JJ quality_NN1 estimates_NN2 ._. 
</s>
<s>
In_RR21 particular_RR22 ,_, since_CS here_RL c=3_FO a_AT1 two-dimensional_JJ plot_NN1 contains_VVZ all_DB the_AT useful_JJ information_NN1 about_II the_AT sample_NN1 ._. 
</s>
<s>
Encouraged_VVN by_II our_APPGE results_NN2 so_RG far_RR ,_, we_PPIS2 prepare_VV0 to_TO implement_VVI all_DB these_DD2 ideas_NN2 in_II a_AT1 prototype_NN1 system_NN1 that_CST would_VM collect_VVI available_JJ Web_NN1 services_NN2 ,_, automatically_RR extract_VV0 metadata_NN1 describing_VVG different_JJ facets_NN2 of_IO these_DD2 services_NN2 and_CC then_RT would_VM use_VVI this_DD1 metadata_NN1 to_TO build_VVI an_AT1 intuitive_JJ search/browse_NN1 interface_NN1 ._. 
</s>
<s>
We_PPIS2 can_VM not_XX ignore_VVI the_AT corpus_NN1 of_IO research_NN1 from_II the_AT social_JJ sciences_NN2 that_CST illustrates_VVZ the_AT unreliability_NN1 of_IO human_JJ perception_NN1 and_CC judgment_NN1 ._. 
</s>
<s>
Each_DD1 AS_RG also_RR announces_VVZ the_AT prefixes_NN2 that_CST it_PPH1 learns_VVZ from_II each_DD1 of_IO its_APPGE neighbors_NN2 to_II its_APPGE other_JJ neighbors_NN2 ._. 
</s>
<s>
Finally_RR ,_, the_AT rekeying_JJ cost_NN1 is_VBZ nine_MC messages_NN2 when_RRQ U1_FO or_CC U2_NP1 departs_VVZ since_CS K1_FO ,_, K2_NP1 ,_, K4_FO ,_, K8_FO and_CC K12_FO need_VV0 to_TO be_VBI changed_VVN ._. 
</s>
<s>
It_PPH1 should_VM be_VBI noted_VVN that_CST ,_, as_CSA the_AT ontology_NN1 ,_, the_AT taxonomy_NN1 of_IO the_AT rules_NN2 could_VM evolve_VVI in_II time_NNT1 to_TO gather_VVI changes_NN2 in_II the_AT ETS_NP2 domain_NN1 and_CC new_JJ requirements_NN2 from_II stakeholders_NN2 and_CC domain_NN1 experts_NN2 ._. 
</s>
<s>
If_CS the_AT decision_NN1 making_NN1 of_IO certain_JJ agents_NN2 is_VBZ influenced_VVN by_II their_APPGE beliefs_NN2 ,_, how_RRQ are_VBR these_DD2 beliefs_NN2 formed_VVN ?_? 
</s>
<s>
The_AT first_MD phase_NN1 of_IO experimentation_NN1 maps_NN2 directly_RR to_II the_AT "_" mine_NN1 "_" and_CC "_" assimilate_VV0 "_" phases_NN2 of_IO the_AT KDD_NP1 process_NN1 shown_VVN in_II Fig._NN1 3_MC ,_, where_CS the_AT data_NN mining_NN1 is_VBZ conducted_VVN to_TO determine_VVI a_AT1 build_NN1 prediction_NN1 model_NN1 using_VVG the_AT Hoeffding_JJ tree_NN1 &lsqb;_( 41_MC &rsqb;_) approach_NN1 and_CC a_AT1 validation_NN1 of_IO the_AT model_NN1 is_VBZ conducted_VVN by_II varying_VVG the_AT order_NN1 in_II which_DDQ the_AT data_NN is_VBZ provided_VVN to_II the_AT classifier_NN1 ._. 
</s>
<s>
Our_APPGE recommendations_NN2 encompass_VV0 many_DA2 interpretations_NN2 of_IO user_NN1 actions_NN2 and_CC numerous_JJ videos_NN2 that_CST users_NN2 may_VM not_XX have_VHI seen_VVN using_VVG normal_JJ query_NN1 methods_NN2 ._. 
</s>
<s>
Adding_VVG the_AT time_NNT1 required_VVN to_TO perform_VVI all_DB the_AT 8000_MC tests_NN2 ,_, the_AT traditional_JJ method_NN1 required_VVN 128665.60_MC seconds_NNT2 ,_, and_CC the_AT CP_NP1 implementations_NN2 required_VVN between_II 27_MC to_II 28_MC seconds_NNT2 (_( no_AT statistical_JJ significant_JJ difference_NN1 was_VBDZ observed_VVN in_II this_DD1 experiment_NN1 between_II using_VVG or_CC not_XX using_VVG symmetry_NN1 breaking_VVG constraints_NN2 )_) ._. 
</s>
<s>
Consumer_NN1 surplus_NN1 can_VM be_VBI quite_RG high_JJ for_IF households_NN2 for_IF whom_PNQO prices_NN2 are_VBR set_VVN considerably_RR below_II the_AT costs_NN2 of_IO supply_NN1 for_IF equity_NN1 reasons_NN2 ._. 
</s>
<s>
BYOD_NP1 Security_NN1 Framework_NN1 implementation_NN1 for_IF an_AT1 enterprise_NN1 reduces_VVZ security_NN1 breaches_NN2 related_VVN to_II lost/stolen_JJ mobile_JJ devices_NN2 ._. 
</s>
<s>
Conjoining_VVG the_AT symmetry_NN1 breaking_VVG predicate_NN1 with_IW the_AT formula_NN1 ensures_VVZ that_CST the_AT SAT_VVD solver_NN1 finds_VVZ few_DA2 representative_JJ assignment_NN1 for_IF every_AT1 equivalence_NN1 class_NN1 ._. 
</s>
<s>
Intra-provider_JJ communication_NN1 is_VBZ based_VVN on_II SMTP_NP1 and_CC a_AT1 secure_JJ TLS_NP1 connection_NN1 ._. 
</s>
<s>
Once_CS the_AT method_NN1 for_IF ensuring_VVG system_NN1 security_NN1 has_VHZ been_VBN set_VVN ,_, the_AT metrics_NN1 should_VM be_VBI established_VVN ,_, and_CC any_DD implementation_NN1 that_CST can_VM not_XX achieve_VVI a_AT1 100%_NNU metric_JJ simply_RR does_VDZ not_XX meet_VVI the_AT design_NN1 goals_NN2 ,_, that_REX21 is_REX22 ,_, it_PPH1 fails_VVZ validation_NN1 ._. 
</s>
<s>
The_AT larger_JJR spreads_NN2 in_II CED_NP1 (_( 104%_NNU )_) ,_, GWP_NP1 (_( 78%_NNU )_) and_CC ADP_NN1 (_( 126%_NNU )_) in_II the_AT recycling_NN1 LCA_NN1 are_VBR mainly_RR higher_JJR due_II21 to_II22 the_AT crediting_JJ alternatives_NN2 of_IO the_AT PS_NN1 ._. 
</s>
<s>
There_EX are_VBR at_RR21 least_RR22 two_MC plausible_JJ reasons_NN2 for_IF this_DD1 gap_NN1 ._. 
</s>
<s>
The_AT Big_JJ Five_MC personality_NN1 model_NN1 was_VBDZ developed_VVN from_II the_AT theoretical_JJ stance_NN1 that_CST personality_NN1 is_VBZ encoded_VVN in_II natural_JJ language_NN1 ,_, and_CC differences_NN2 in_II personality_NN1 may_VM become_VVI apparent_JJ through_II linguistic_JJ variations_NN2 &lsqb;_( 32_MC &rsqb;_) ._. 
</s>
<s>
If_CS we_PPIS2 need_VV0 to_TO make_VVI these_DD2 two_MC types_NN2 of_IO models_NN2 work_VV0 in_II concert_NN1 as_CSA components_NN2 ,_, we_PPIS2 find_VV0 that_CST the_AT complexity_NN1 of_IO these_DD2 components_NN2 is_VBZ in_II different_JJ dimensions_NN2 ,_, which_DDQ would_VM be_VBI hard_JJ to_TO compare_VVI ._. 
</s>
<s>
Using_VVG the_AT above_JJ notations_NN2 we_PPIS2 present_VV0 the_AT solution_NN1 of_IO almost-sure_JJ winning_VVG for_IF safety_NN1 and_CC reachability_NN1 objectives_NN2 ._. 
</s>
<s>
Rigoutsos_NN2 and_CC Huynh_NP1 (_( 2004_MC )_) apply_VV0 the_AT Teiresias_NP2 pattern_NN1 discovery_NN1 algorithm_NN1 to_TO email_VVI classification_NN1 ._. 
</s>
<s>
We_PPIS2 then_RT present_VV0 the_AT user_NN1 interface_NN1 built_VVN on_II31 top_II32 of_II33 integrated_JJ datasets_NN2 as_II31 well_II32 as_II33 outline_NN1 various_JJ features_NN2 of_IO our_APPGE interface_NN1 ._. 
</s>
<s>
The_AT handoff_NN1 locator_NN1 provides_VVZ an_AT1 explicit_JJ in-direction_JJ mechanism_NN1 to_TO determine_VVI the_AT ingress_NN1 of_IO each_DD1 policy_NN1 domain_NN1 ._. 
</s>
<s>
Behavior_NN1 specification_NN1 can_VM be_VBI structured_VVN using_VVG components_NN2 that_CST abstract_VV0 away_RL message_NN1 and_CC data_NN details_NN2 ._. 
</s>
<s>
In_II the_AT limit_NN1 of_IO very_RG large_JJ values_NN2 for_IF β_NULL β_NULL ,_, hh_NNU tends_VVZ toward_II the_AT value_NN1 of_IO the_AT coercivity_NN1 constant_NN1 for_IF a_AT1 piecewise_JJ linear_JJ ,_, conforming_VVG finite_JJ element_NN1 approximation_NN1 on_II the_AT same_DA mesh_NN1 (_( the_AT conforming_VVG curve_NN1 is_VBZ not_XX shown_VVN )_) ._. 
</s>
<s>
Here_RL ,_, we_PPIS2 assume_VV0 that_CST the_AT images_NN2 to_TO be_VBI segmented_VVN have_VH0 no_AT object_NN1 overlapping_VVG with_IW the_AT outer_JJ most_DAT rows_NN2 and_CC columns_NN2 of_IO the_AT image_NN1 ._. 
</s>
<s>
However_RR ,_, the_AT gains_NN2 and_CC benefits_NN2 well_RR outweigh_VV0 the_AT costs_NN2 (_( Ramesh_NP1 ,_, 1998_MC )_) ._. 
</s>
<s>
A_AT1 descriptor_NN1 with_IW more_RGR detailed_JJ shape_NN1 information_NN1 has_VHZ been_VBN proposed_VVN by_II Belongie_NP1 et_RA21 al_RA22 ._. 
</s>
<s>
&lsqb;_( 40_MC &rsqb;_) ,_, where_CS each_DD1 point_NN1 on_II the_AT contour_NN1 is_VBZ associated_VVN with_IW a_AT1 shape_NN1 context_NN1 describing_VVG the_AT coarse_JJ arrangement_NN1 of_IO the_AT rest_NN1 of_IO the_AT shape_NN1 with_II31 respect_II32 to_II33 the_AT point_NN1 ._. 
</s>
<s>
In_II31 terms_II32 of_II33 query_NN1 resolution_NN1 delay_NN1 and_CC QDR_NP1 we_PPIS2 observe_VV0 that_CST the_AT increase_NN1 in_II the_AT size_NN1 of_IO the_AT network_NN1 does_VDZ not_XX impact_NN1 significantly_RR on_II the_AT performance_NN1 of_IO both_DB2 the_AT double_JJ ruling_JJ scheme_NN1 and_CC the_AT crossroads_NN1 approach_NN1 ._. 
</s>
<s>
Fig._NN1 14presents_FO plots_NN2 for_IF the_AT penalty_NN1 incurred_VVN and_CC the_AT number_NN1 of_IO violations_NN2 prevalent_JJ ,_, along_II21 with_II22 the_AT 95%_NNU confidence_NN1 interval_NN1 for_IF each_DD1 value_NN1 ,_, when_CS native_JJ transit_NN1 policy_NN1 violations_NN2 are_VBR disallowed_VVN by_II a_AT1 certain_JJ number_NN1 of_IO host_NN1 ASes_NP2 ._. 
</s>
<s>
We_PPIS2 experimented_VVD both_DB2 alternatives_NN2 and_CC we_PPIS2 will_VV0 shown_VVN in_II Section_NN1 6.2_MC that_DD1 ,_, the_AT choice_NN1 of_IO any_DD of_IO these_DD2 fitness_NN1 functions_NN2 leads_VVZ to_II equivalent_JJ results_NN2 both_RR in_II31 terms_II32 of_II33 the_AT number_NN1 of_IO features_NN2 selected_VVN and_CC performance_NN1 ._. 
</s>
<s>
If_CS all_DB four_MC policy_NN1 levels_NN2 fully_RR trust_NN1 signed_VVD Microsoft_NP1 assemblies_NN2 ,_, then_RT any_DD assembly_NN1 from_II Microsoft_NP1 is_VBZ fully_RR trusted_VVN on_II that_DD1 machine_NN1 ._. 
</s>
<s>
The_AT key_JJ components_NN2 of_IO the_AT EEM_NN1 program_NN1 include_VV0 a_AT1 fish_NN survey_VV0 to_TO assess_VVI effects_NN2 on_II fish_NN and_CC a_AT1 benthic_JJ invertebrate_JJ community_NN1 survey_NN1 to_TO assess_VVI effects_NN2 on_II fish_NN habitat_NN1 ._. 
</s>
<s>
Clusters_NN2 which_DDQ are_VBR not_XX re-populated_VVN with_IW any_DD FQDNs_NP1 during_II ?_? δMa_FO ?_? are_VBR removed_VVN ,_, and_CC therefore_RR "_" live_VV0 "_" longer_JJR thanδMa_FO ._. 
</s>
<s>
This_DD1 point_NN1 is_VBZ emphasized_VVN in_II the_AT model_NN1 of_IO this_DD1 study_NN1 by_II including_VVG a_AT1 semi-circular_JJ arrow_NN1 in_II the_AT upper_JJ left_JJ corner_NN1 of_IO the_AT process_NN1 ._. 
</s>
<s>
For_IF Plots_NN2 Fig._NN1 4_NN1 (_( a_AT1 and_CC c_ZZ1 )_) we_PPIS2 assume_VV0 that_CST there_EX are_VBR no_AT storage_NN1 costs_NN2 ,_, whereas_CS for_IF Plots_NN2 Fig._NN1 4_NN1 (_( b_ZZ1 and_CC d_ZZ1 )_) the_AT storage_NN1 cost_NN1 rate_NN1 is_VBZ set_VVN to_II 0_MC for_IF nodes_NN2 with_IW odd_JJ index_NN1 ,_, and_CC equal_JJ to_II for_IF nodes_NN2 with_IW even_JJ index_NN1 ._. 
</s>
<s>
Finally_RR ,_, the_AT Hardware_NN1 Trojan_JJ opens_VVZ a_AT1 back_NN1 door_NN1 to_II the_AT network_NN1 endpoint_NN1 and_CC suspends_VVZ its_APPGE attack_NN1 by_II returning_VVG the_AT network_NN1 endpoint_NN1 to_II the_AT state_NN1 in_II which_DDQ the_AT authorized_JJ user_NN1 had_VHD left_VVN it_PPH1 ._. 
</s>
<s>
Nonetheless_RR ,_, barriers_NN2 to_II accessing_NN1 and_CC using_VVG complex_JJ information_NN1 for_IF supporting_VVG regional_JJ policy_NN1 ,_, planning_VVG ,_, and_CC management_NN1 decisions_NN2 still_RR remain_VV0 due_II21 to_II22 several_DA2 factors_NN2 including_II limited_JJ quantitative_JJ analytical_JJ capacity_NN1 ,_, limited_JJ time_NNT1 and_CC willingness_NN1 to_TO engage_VVI with_IW information_NN1 ,_, and_CC the_AT need_NN1 for_IF specialized_JJ technical_JJ requirements_NN2 such_II21 as_II22 software_NN1 or_CC data_NN ._. 
</s>
<s>
The_AT transition_NN1 3_MC →_NULL 4_MC is_VBZ based_VVN on_II this_DD1 probability_NN1 which_DDQ is_VBZ represented_VVN by_II P34_FO in_II the_AT model_NN1 ._. 
</s>
<s>
After_CS the_AT 12th_MD second_NNT1 ,_, the_AT node_NN1 will_VM have_VHI moved_VVN into_II final_JJ position_NN1 and_CC it_PPH1 '_NULL s_ZZ1 edges_NN2 become_VV0 dark_JJ blue_JJ ._. 
</s>
<s>
These_DD2 additional_JJ virtual_JJ packets_NN2 are_VBR then_RT stored_VVN in_II the_AT corresponding_JJ queue_NN1 and_CC will_VM receive_VVI service_NN1 at_II the_AT flow_NN1 '_NULL s_ZZ1 guaranteed_JJ rate_NN1 ._. 
</s>
<s>
Some_DD of_IO these_DD2 were_VBDR clearly_RR indicative_JJ of_IO support_NN1 for_IF the_AT leave/remain_VV0 campaigns_NN2 ,_, e.g._REX #votetoleave_FO ,_, #voteout_FO ,_, #saferin_FO ,_, #strongertogether_FO ._. 
</s>
<s>
GIS_NN2 practitioners_NN2 use_VV0 geospatial_JJ relational_JJ databases_NN2 in_II their_APPGE day-to-day_JJ tasks_NN2 ,_, either_RR directly_RR or_CC as_II the_AT back-end_NN1 of_IO applications_NN2 to_TO store_VVI and_CC manipulate_VVI data_NN (_( e.g._REX ,_, GIS_NN2 have_VH0 connectors_NN2 for_IF geospatial_JJ relational_JJ databases_NN2 )_) ._. 
</s>
<s>
Not_XX surprisingly_RR ,_, the_AT authors_NN2 report_VV0 a_AT1 weak_JJ performance_NN1 of_IO these_DD2 taggers_NN2 when_CS applied_VVN to_II tweets_NN2 ._. 
</s>
<s>
Each_DD1 of_IO these_DD2 concatenated_JJ labels_NN2 is_VBZ mapped_VVN and_CC replaced_VVN with_IW a_AT1 unique_JJ new_JJ label_NN1 (_( step_NN1 3_MC and_CC 4_MC )_) ,_, as_CSA in_II the_AT Weisfeiler–Lehman_NN1 algorithm_NN1 ._. 
</s>
<s>
We_PPIS2 used_VVD the_AT following_JJ settings_NN2 :_: K=3_FO ,_, p=3_FO ,_, k=13_FO (_( k=1_FO ,_, ,_, 3_MC )_) ,_, 1=_FO (_( 0,0,1_MC )_) T_ZZ1 ,_, 2=_FO (_( 0,1,0_MC )_) T_ZZ1 ,_, 3=_FO (_( 1,0,0_MC )_) T._NP1 Another_DD1 parameter_NN1 is_VBZ variance-covariance_JJ matrix_NN1 ._. 
</s>
<s>
Let_VV0 dist_NN1 (_( v_ZZ1 ,_, v_ZZ1 )_) define_VV0 the_AT length_NN1 of_IO the_AT shortest_JJT path_NN1 in_II G_ZZ1 between_II the_AT vertices_NN2 v_ZZ1 and_CC v_ZZ1 ,_, and_CC let_VV0 G=_FO (_( V_ZZ1 ,_, E_ZZ1 ,_, )_) be_VBI an_AT1 RDF_NP1 graph_NN1 ._. 
</s>
<s>
Sections_NN2 4_MC and_CC 5_MC detail_VV0 the_AT image_NN1 sets_VVZ that_CST have_VH0 been_VBN used_VVN and_CC the_AT protocols_NN2 followed_VVN for_IF off-line_JJ and_CC on-line_JJ testing_NN1 ._. 
</s>
<s>
Some_DD consider_VV0 a_AT1 disease_NN1 to_TO be_VBI a_AT1 process_NN1 ,_, and_CC hence_RR an_AT1 occurrent_NN1 &lsqb;_( 12_MC &rsqb;_) ._. 
</s>
<s>
In_BCL21 order_BCL22 to_TO find_VVI a_AT1 starting_NN1 point_NN1 for_IF minimization_NN1 of_IO the_AT function_NN1 fk_NNU ,_, in_II Step_NN1 1_MC1 we_PPIS2 first_MD determine_VV0 all_DB data_NN points_NN2 which_DDQ attract_VV0 at_RR21 least_RR22 one_MC1 data_NN point_NN1 ._. 
</s>
<s>
Thus_RR ,_, our_APPGE implementation_NN1 takes_VVZ as_CSA input_NN1 application_NN1 codeP_NN1 ,_, a_AT1 secret_JJ key_NN1 ,_, and_CC a_AT1 watermark_NN1 value_NN1 ._. 
</s>
<s>
CUDA_NP1 also_RR exposes_VVZ the_AT internal_JJ architecture_NN1 of_IO the_AT GPU_NP1 and_CC allows_VVZ direct_JJ access_NN1 to_II its_APPGE internal_JJ resources_NN2 ._. 
</s>
<s>
The_AT figures_NN2 show_VV0 that_CST the_AT bias_NN1 decreases_VVZ withrfor_NN1 ther-fold_JJ estimators_NN2 ._. 
</s>
<s>
UCMs_NP1 provide_VV0 benefits_NN2 similar_JJ to_II those_DD2 cited_VVN by_II Binder_NN1 ,_, but_CCB they_PPHS2 also_RR provide_VV0 an_AT1 appropriate_JJ level_NN1 of_IO abstraction_NN1 for_IF early_JJ design_NN1 stages_NN2 ._. 
</s>
<s>
We_PPIS2 need_VV0 studies_NN2 to_TO address_VVI this_DD1 issue_NN1 ._. 
</s>
<s>
Let_VV0 V_ZZ1 be_VBI either_RR the_AT least_RRT leaf_VV0 value_NN1 greater_JJR than_CSN V_ZZ1 or_CC any_DD value_NN1 greater_JJR than_CSN the_AT max_NN1 leaf_NN1 if_CS V_ZZ1 is_VBZ the_AT maximum_NN1 ._. 
</s>
<s>
While_CS the_AT algorithms_NN2 implemented_VVN do_VD0 not_XX meet_VVI the_AT formalism_NN1 of_IO CA_NP1 ,_, they_PPHS2 do_VD0 share_VVI several_DA2 key_JJ characteristics_NN2 ._. 
</s>
<s>
However_RR ,_, given_CS21 that_CS22 a_AT1 compromised_JJ primary_JJ master_NN1 can_VM implement_VVI more_RGR effective_JJ attacks_NN2 on_II the_AT power_NN1 system_NN1 ,_, e.g._REX by_II directly_RR disconnecting_VVG the_AT residential_JJ home_NN1 from_II the_AT grid_NN1 ,_, accessing_VVG consumer_NN1 only_JJ functionality_NN1 would_VM not_XX be_VBI a_AT1 priority_NN1 target_NN1 ._. 
</s>
<s>
We_PPIS2 chose_VVD PubMed_JJ because_CS it_PPH1 contains_VVZ more_DAR than_CSN 23_MC million_NNO publications_NN2 and_CC provides_VVZ an_AT1 interface_NN1 that_CST allows_VVZ discovering_VVG novel_JJ publications_NN2 as_CS31 soon_CS32 as_CS33 they_PPHS2 are_VBR made_VVN available_JJ ._. 
</s>
<s>
Yet_RR even_RR with_IW this_DD1 suggestion_NN1 problems_NN2 remain_VV0 ,_, including_VVG that_CST people_NN do_VD0 not_XX always_RR follow_VVI security_NN1 policies_NN2 (_( &lsqb;_( Workman_NP1 ,_, 2008a_FO &rsqb;_) and_CC &lsqb;_( Workman_NP1 ,_, 2008b_FO &rsqb;_) )_) ,_, ethics_NN training_NN1 is_VBZ not_XX always_RR effective_JJ in_II addressing_VVG security-related_JJ behavior_NN1 (_( Workman_NP1 and_CC Gathegi_NP1 ,_, 2007_MC )_) ,_, and@there_FO is_VBZ little_JJ in_II the_AT literature_NN1 about_II what_DDQ factors_NN2 lead_VV0 to_II cyber_NN1 harassment_NN1 against_II corporations_NN2 ._. 
</s>
<s>
While_CS this_DD1 might_VM be_VBI possible_JJ for_IF readers_NN2 ,_, it_PPH1 places_VVZ severe_JJ storage_NN1 constraints_NN2 on_II the_AT tags_NN2 ._. 
</s>
<s>
As_II the_AT ontologies_NN2 are_VBR consistent_JJ ,_, the_AT annotations_NN2 also_RR benefit_VV0 from_II this_DD1 consistency_NN1 ._. 
</s>
<s>
Even_CS21 though_CS22 the_AT proposed_JJ framework_NN1 provides_VVZ a_AT1 means_NN for_IF analyzing_VVG false_JJ positive_JJ alerts_NN2 (_( roughly_RR speaking_VVG the_AT attack_NN1 probability_NN1 of_IO patterns_NN2 will_VM increase_VVI with_IW their_APPGE lengths_NN2 )_) ,_, the_AT stored_JJ pattern_NN1 tree_NN1 might_VM include_VVI some_DD false_JJ positive_JJ alerts_NN2 (_( the_AT leaf_NN1 nodes_NN2 with_IW depth_NN1 1_MC1 ,_, in_RR21 particular_RR22 )_) ._. 
</s>
<s>
The_AT second_MD strategy_NN1 aims_VVZ at_II detecting_VVG the_AT error_NN1 and_CC stopping_VVG it_PPH1 before_II damages_NN2 occur_VV0 ,_, while_CS the_AT last_MD strategy_NN1 is_VBZ intended_VVN at_II handling_VVG errors_NN2 that_CST have_VH0 already_RR occurred_VVN and_CC providing_VVG some_DD form_NN1 of_IO reversibility_NN1 ._. 
</s>
<s>
During_II the_AT expansion_NN1 phase_NN1 ,_, pseudo-nodes_NN2 are_VBR replaced_VVN with_IW the_AT original_JJ cycles_NN2 ._. 
</s>
<s>
In_RR21 particular_RR22 ,_, the_AT provenance_NN1 of_IO information_NN1 is_VBZ crucial_JJ in_II deciding_VVG whether_CSW information_NN1 is_VBZ to_TO be_VBI trusted_VVN ,_, how_RRQ it_PPH1 should_VM be_VBI integrated_VVN with_IW other_JJ diverse_JJ information_NN1 sources_NN2 ,_, and_CC how_RRQ to_TO give_VVI credit_NN1 to_II its_APPGE originators_NN2 when_CS reusing_VVG it_PPH1 ._. 
</s>
<s>
Aside_II21 from_II22 previous_JJ practices_NN2 ,_, the_AT resource_NN1 state_NN1 design_NN1 in_II IEM_NP1 may_VM face_VVI specific_JJ requirements_NN2 ,_, such_II21 as_II22 time-based_JJ execution_NN1 of_IO models_NN2 and_CC the_AT ability_NN1 to_TO step_VVI back_RP to_II a_AT1 previous_JJ state_NN1 of_IO a_AT1 given_JJ resource_NN1 ._. 
</s>
<s>
Solutions_NN2 from_II this_DD1 strategy_NN1 are_VBR usually_RR not_XX expected_VVN to_TO be_VBI the_AT most_RGT optimal_JJ ones_NN2 overall_RR when_CS all_DB the_AT RT_NN1 violations_NN2 are_VBR considered_VVN in_II a_AT1 holistic_JJ approach_NN1 using_VVG global_JJ optimization_NN1 techniques_NN2 ._. 
</s>
<s>
Step_NN1 1_MC1 :_: Inspecting_VVG SRS_NP1 for_IF faults_NN2 using_VVG the_AT fault_NN1 checklist_NN1 :_: Using_VVG information_NN1 from_II Training_NN1 1_MC1 ,_, each_DD1 subject_NN1 inspected_VVD the_AT requirement_NN1 document_NN1 using_VVG a_AT1 fault_NN1 checklist_NN1 ._. 
</s>
<s>
For_IF the_AT -maximal_JJ set_NN1 A_ZZ1 Sadmissible_JJ in_II (_( X_ZZ1 ,_, A_ZZ1 )_) ,_, it_PPH1 holds_VVZ that_DD1 ._. 
</s>
<s>
The_AT intent_NN1 behind_II such_DA attacks_NN2 can_VM be_VBI narrowed_VVN down_RP to_II a_AT1 handful_NN1 of_IO possibilities_NN2 ._. 
</s>
<s>
The_AT Yale_NP1 B_ZZ1 database_NN1 consists_VVZ of_IO 5850_MC facial_JJ images_NN2 of_IO 10_MC persons_NN2 ,_, which_DDQ has_VHZ 585_MC images_NN2 per_II person_NN1 having_VHG nine_MC poses_VVZ under_RG 65_MC illumination_NN1 conditions_NN2 with_IW an_AT1 image_NN1 resolution_NN1 of_IO 480×640_FO pixels_NN2 ._. 
</s>
<s>
This_DD1 paper_NN1 demonstrates_VVZ that_DD1 information_NN1 assurance_NN1 awareness_NN1 and_CC training_NN1 can_VM be_VBI provided_VVN in_II an_AT1 engaging_JJ format_NN1 ._. 
</s>
<s>
These_DD2 images_NN2 have_VH0 been_VBN chosen_VVN based_VVN on_II some_DD of_IO the_AT major_JJ applications_NN2 of_IO skeletonisation_NN1 ,_, and_CC the_AT image_NN1 processing_NN1 field_NN1 in_RR21 general_RR22 ._. 
</s>
<s>
Pioneer_NN1 2AT_FO is_VBZ a_AT1 four-wheeled_JJ robot_NN1 that_CST is_VBZ equipped_VVN with_IW a_AT1 differential_JJ drive_NN1 train_NN1 system_NN1 and_CC has_VHZ an_AT1 approximate_JJ weight_NN1 of_IO 35_MC kg_NNU ._. 
</s>
<s>
The_AT object_NN1 sizes_NN2 follow_VV0 a_AT1 Pareto_NP1 distribution_NN1 with_IW shape_NN1 parameter_NN1 1.2_MC ,_, and_CC the_AT inter-page_NN1 and_CC inter-object_JJ time_NNT1 distributions_NN2 were_VBDR exponential_NN1 with_IW means_NN 9_MC s_ZZ1 and_CC 1_MC1 ms_NN1 respectively_RR ._. 
</s>
<s>
Device_NN1 Count_NN1 :_: The_AT number_NN1 of_IO firewall_NN1 devices_NN2 in_II operation_NN1 ._. 
</s>
<s>
For_IF the_AT sake_NN1 of_IO brevity_NN1 ,_, the_AT definition_NN1 of_IO setbnd_NN1 published_VVN in_II &lsqb;_( 17_MC &rsqb;_) assumed_VVD a_AT1 one-cell_JJ border_NN1 of_IO solid_JJ cells_NN2 on_II the_AT outside_II21 of_II22 the_AT cell-space_NN1 ._. 
</s>
<s>
To_TO understand_VVI the_AT global_JJ behavior_NN1 of_IO the_AT model_NN1 and_CC the_AT impact_NN1 of_IO both_DB2 and_CC β_NULL ,_, we_PPIS2 constructed_VVD a_AT1 phase_NN1 diagram_NN1 for_IF ranges_NN2 of_IO and_CC β_NULL (_( see_VV0 Fig._NN1 8_MC )_) ._. 
</s>
<s>
Step_NN1 one_MC1 was_VBDZ the_AT identification_NN1 of_IO the_AT key_JJ variables_NN2 of_IO interest_NN1 to_II the_AT managers_NN2 ._. 
</s>
<s>
We_PPIS2 also_RR describe_VV0 several_DA2 heuristics_NN2 and_CC health-related_JJ models_NN2 that_CST have_VH0 potential_JJ for_IF improving_JJ cyber_NN1 security_NN1 ._. 
</s>
<s>
Additionally_RR ,_, the_AT validation_NN1 measures_NN2 in_II STAC_NP1 allow_VV0 two_MC systems_NN2 of_IO different_JJ design_NN1 with_IW similar_JJ missions_NN2 to_TO be_VBI compared_VVN to_II see_VV0 if_CSW one_PN1 is_VBZ architecture_NN1 is_VBZ more_RGR secure_JJ than_CSN the_AT other_JJ ._. 
</s>
<s>
This_DD1 is_VBZ likened_VVN to_II the_AT risk_NN1 of_IO such_DA a_AT1 thing_NN1 to_TO happen_VVI ._. 
</s>
<s>
The_AT aforementioned_JJ state_NN1 evolution_NN1 changes_NN2 upon_II reception_NN1 of_IO a_AT1 notification_NN1 that_CST signals_VVZ the_AT occurrence_NN1 of_IO a_AT1 Multiple_JJ Resource_NN1 Failure_NN1 ._. 
</s>
<s>
Present_JJ recruitment_NN1 is_VBZ less_DAR than_CSN 1/10_MF of_IO its_APPGE historical_JJ level_NN1 (_( Dekker_NP1 ,_, 2003_MC ,_, ICES_NN2 ,_, 2012_MC )_) and_CC the_AT species_NN is_VBZ considered_VVN critically_RR endangered_VVN (_( UN_NP1 CITES_VVZ Appendix_NN1 II_MC ,_, IUCN_VV0 Red_JJ List_NN1 )_) ._. 
</s>
<s>
The_AT data_NN import_VV0 task_NN1 is_VBZ also_RR prone_JJ to_II errors_NN2 ._. 
</s>
<s>
Recently_RR ,_, there_EX has_VHZ been_VBN some_DD interesting_JJ debate_NN1 with_II31 respect_II32 to_II33 the_AT validity_NN1 of_IO Metcalfe_NP1 '_NULL s_ZZ1 law_NN1 ._. 
</s>
<s>
Utility_NN1 companies_NN2 have_VH0 not_XX been_VBN amenable_JJ in_II adapting_VVG to_II the_AT changing_JJ demands_NN2 in_II communication_NN1 networks_NN2 to_TO support_VVI increasing_JJ smart_JJ grid_NN1 tools_NN2 and_CC applications_NN2 for_IF several_DA2 reasons_NN2 ._. 
</s>
<s>
In_II its_APPGE 2008_MC fourth_MD edition_NN1 ,_, the_AT PMBOK_NN1 Guide_NN1 also_RR put_VV0 more_DAR emphasis_NN1 on_II the_AT role_NN1 of_IO project_NN1 stakeholders_NN2 and_CC managing_VVG their_APPGE expectations_NN2 as_CSA success_NN1 factors_NN2 for_IF PM_NP1 ,_, though_CS not_XX as_RG much_DA1 emphasis_NN1 as_CSA it_PPH1 put_VVD on_II user_NN1 and_CC management_NN1 issues_NN2 &lsqb;_( 29_MC &rsqb;_) ._. 
</s>
<s>
The_AT meaning_NN1 of_IO the_AT term_NN1 "_" cubicle_NN1 ,_, "_" for_REX21 example_REX22 ,_, was_VBDZ completely_RR different_JJ at_II the_AT mid-US_NP1 site_NN1 and_CC at_II the_AT India_NP1 site_NN1 ,_, due_II21 to_II22 local_JJ differences_NN2 such_II21 as_II22 the_AT height_NN1 of_IO cubicle_NN1 dividers_NN2 ,_, the_AT number_NN1 of_IO people_NN per_II cubicle_NN1 ,_, the_AT area_NN1 of_IO the_AT cubicle_NN1 ,_, etc_RA ._. 
</s>
<s>
The_AT remaining_JJ values_NN2 of_IO the_AT parameters_NN2 used_VVN in_II calculations_NN2 are_VBR tabulated_VVN inTable_JJ 4_MC ._. 
</s>
<s>
Split_JJ plan_NN1 processing_NN1 incurs_VVZ significant_JJ deployment_NN1 latency_NN1 ._. 
</s>
<s>
Descriptive_JJ statistics_NN (_( mean_VV0 and_CC standard_JJ deviations_NN2 )_) and_CC correlations_NN2 (_( Pearson_NP1 product-moment_NN1 )_) for_IF the_AT study_NN1 '_NULL s_ZZ1 variables_NN2 are_VBR displayed_VVN in_II Table_NN1 2_MC ._. 
</s>
<s>
In_II this_DD1 Section_NN1 we_PPIS2 first_MD develop_VV0 a_AT1 principled_JJ axiomatisation_NN1 of_IO the_AT observability_NN1 of_IO actions_NN2 ,_, then_RT build_VV0 a_AT1 powerful_JJ yet_RR succinct_JJ axiomatisation_NN1 of_IO knowledge_NN1 upon_II it_PPH1 ._. 
</s>
<s>
The_AT keyboard_NN1 used_VVN in_II this_DD1 study_NN1 is_VBZ very_RG small_JJ ,_, with_IW a_AT1 more_RGR restricted_JJ keystroke_NN1 interface_NN1 thus_RR reduced_VVN distance_NN1 between_II the_AT keys_NN2 ._. 
</s>
<s>
The_AT hypergraph_NN1 ranking_NN1 algorithm_NN1 is_VBZ the_AT same_DA as_CSA Algorithm_NN1 1_MC1 except_II21 for_II22 using_VVG the_AT binary_JJ incidence_NN1 matrix_NN1 (_( where_CS ht_NNU (_( vi_MC ,_, ej_NN1 )_) =1_FO or_CC 0_MC )_) ._. 
</s>
<s>
The_AT estimated_JJ frequencies_NN2 were_VBDR obtained_VVN for_IF each_DD1 variable_NN1 ,_, using_VVG the_AT periodogram_NN1 to_TO preserve_VVI the_AT linearity_NN1 of_IO the_AT model_NN1 with_II31 respect_II32 to_II33 the_AT unknown_JJ parameters_NN2 (_( Table_NN1 4_MC )_) ._. 
</s>
<s>
The_AT training_NN1 was_VBDZ same_DA as_CSA in_II Experiment_NN1 1_MC1 except_CS21 that_CS22 the_AT subjects_NN2 were_VBDR trained_VVN how_RRQ to_TO use_VVI the_AT revised_JJ NLtoSTD-BB_NN1 V2.0_FO method_NN1 to_TO locate_VVI faults_NN2 using_VVG examples_NN2 and_CC a_AT1 practice_NN1 problem_NN1 ,_, and_CC how_RRQ to_TO record_VVI faults_NN2 ._. 
</s>
<s>
As_II a_AT1 consequence_NN1 of_IO missing_JJ messages_NN2 ,_, we_PPIS2 recognize_VV0 that_CST some_DD users_NN2 will_VM appear_VVI to_TO remain_VVI unconnected_JJ or_CC connected_VVN by_II a_AT1 path_NN1 of_IO longer_JJR length_NN1 ._. 
</s>
<s>
Another_DD1 new_JJ feature_NN1 is_VBZ the_AT addition_NN1 of_IO the_AT '_NULL Trust_NN1 Center_NN1 '_NULL ._. 
</s>
<s>
Specific_JJ applications_NN2 of_IO LSP_NP1 information_NN1 include_VV0 for_REX21 example_REX22 the_AT quantification_NN1 of_IO crop_NN1 yields_NN2 ,_, wildfire_NN1 fuel_NN1 accumulation_NN1 ,_, ecosystem_VV0 resilience_NN1 and_CC health_NN1 and_CC land_NN1 surface_NN1 modelling_NN1 (_( Eklundh_NP1 and_CC J_ZZ1 ?_? nsson_NN1 ,_, 2010_MC ,_, Friedl_NP1 et_NP1 ?_? al._NNU ,_, 2006_MC ,_, Liang_NP1 and_CC Schwartz_NP1 ,_, 2009_MC ,_, Pe_NN1 ?_? uelas_NN2 et_NP1 ?_? al._NNU ,_, 2009_MC ;_; Schwartz_NP1 ,_, 2013_MC )_) ._. 
</s>
<s>
Given_VVN a_AT1 SPARQL_NN1 query_NN1 q_ZZ1 ,_, we_PPIS2 denote_VV0 by_II rewgeo(q)the_NN1 SPARQL_NN1 query_NN1 rewritten_VVN by_II Rgeo_NP1 ._. 
</s>
<s>
Given_VVN an_AT1 assignment_NN1 matrix_NN1 ?_? 
</s>
<s>
X_ZZ1 ,_, VM_NP1 requests_NN2 ,_, topology_NN1 ,_, PMs_NP2 ,_, sinks_NN2 and_CC link_VV0 related_JJ information_NN1 the_AT challenge_NN1 is_VBZ to_TO fill_VVI the_AT entries_NN2 of_IO the_AT matrix_NN1 ?_? 
</s>
<s>
X_ZZ1 ?_? with_IW 0s_MC2 and_CC 1s_MC2 so_CS21 that_CS22 it_PPH1 maximizes_VVZ the_AT objective_JJ function_NN1 and_CC does_VDZ not_XX violate_VVI any_DD constraint_NN1 ._. 
</s>
<s>
This_DD1 involves_VVZ constantly_RR monitoring_VVG the_AT incoming_JJ streams_NN2 and_CC making_VVG the_AT appropriate_JJ changes_NN2 on_II join_NN1 orders_NN2 and_CC the_AT related_JJ index_NN1 structures_NN2 whenever_RRQV beneficial_JJ ._. 
</s>
<s>
Since_CS the_AT object_NN1 does_VDZ not_XX occlude_VVI image_NN1 edges_NN2 ,_, the_AT image_NN1 values_NN2 are_VBR already_RR zero_MC towards_II image_NN1 edges_NN2 ,_, and_CC there_EX is_VBZ no_AT need_NN1 for_IF a_AT1 "_" windowing_JJ "_" operation_NN1 ._. 
</s>
<s>
It_PPH1 is_VBZ also_RR possible_JJ to_TO envision_VVI many_DA2 different_JJ scenarios_NN2 where_RRQ a_AT1 Hardware_NN1 Trojan_JJ based_VVN on_II USB_NP1 peripherals_NN2 could_VM be_VBI used_VVN ,_, by_II a_AT1 malicious_JJ insider_NN1 ._. 
</s>
<s>
For_REX21 example_REX22 ,_, increase_VV0 in_II the_AT value_NN1 of_IO security_NN1 may_VM decrease_VVI the_AT value_NN1 of_IO performance_NN1 for_IF the_AT stakeholders_NN2 &lsqb;_( 8_MC &rsqb;_) ._. 
</s>
<s>
Understanding_VVG the_AT performance_NN1 across_II islands_NN2 would_VM require_VVI us_PPIO2 to_TO assess_VVI the_AT latency_NN1 and_CC bandwidth_NN1 characteristics_NN2 of_IO the_AT inter-island_JJ links_NN2 (_( which_DDQ would_VM require_VVI a_AT1 special_JJ access_NN1 mode_NN1 )_) ,_, and_CC incorporate_VV0 these_DD2 in_II a_AT1 separate_JJ "_" inter-island_JJ "_" set_NN1 of_IO the_AT parameters_NN2 and_CC ._. 
</s>
<s>
Each_DD1 sentence_NN1 ends_VVZ with_IW a_AT1 conventional_JJ symbol_NN1 <EOS>_FO that_DD1 indicates_VVZ the_AT end_NN1 of_IO the_AT sentence_NN1 ._. 
</s>
<s>
All_DB of_IO these_DD2 conditions_NN2 led_VVN to_II a_AT1 more_RGR stringent_JJ assessment_NN1 of_IO the_AT results_NN2 ._. 
</s>
<s>
Construction_NN1 of_IO an_AT1 explicit_JJ LSF_NP1 using_VVG SVM_NP1 in_II the_AT space_NN1 of_IO coefficients_NN2 ._. 
</s>
<s>
In_II his_APPGE work_NN1 ,_, Ko_NN1 presented_VVD partial_JJ specifications_NN2 generated_VVN for_IF several_DA2 programs_NN2 in_II31 terms_II32 of_II33 their_APPGE specification_NN1 lengths_NN2 ._. 
</s>
<s>
IEM_VV0 applications_NN2 are_VBR the_AT stakeholder_NN1 community_NN1 '_NULL s_ZZ1 methods_NN2 for_IF selecting_VVG ,_, organizing_VVG ,_, integrating_VVG ,_, and_CC processing_VVG the_AT combination_NN1 of_IO environmental_JJ ,_, social_JJ ,_, and_CC economic_JJ information_NN1 needed_VVD to_TO inform_VVI decisions_NN2 and_CC policies_NN2 related_VVN to_II the_AT environment_NN1 ._. 
</s>
<s>
In_II this_DD1 organization_NN1 ,_, clusters_NN2 of_IO strongly_RR connected_VVN pure_JJ FIL_NNU1 index_NN1 networks_NN2 are_VBR connected_VVN by_II NSL_NP1 search_NN1 links_NN2 ._. 
</s>
<s>
Thus_RR ,_, youtube_VV0 video_NN1 was_VBDZ labeled_VVN as_CSA content_JJ and_CC most_RGT popular_JJ as_CSA intent_JJ here_RL (_( we_PPIS2 believe_VV0 rightly_RR so_RR )_) ._. 
</s>
<s>
While_CS this_DD1 too_RR is_VBZ expected_VVN (_( normal_JJ triangles_NN2 make_VV0 up_RP the_AT vast_JJ majority_NN1 of_IO most_DAT stroke-based_JJ images_NN2 )_) ,_, Table_NN1 2_MC highlights_VVZ another_DD1 issue_NN1 ._. 
</s>
<s>
This_DD1 paper_NN1 is_VBZ an_AT1 extended_JJ version_NN1 of_IO a_AT1 preliminary_JJ conference_NN1 paper_NN1 &lsqb;_( 22_MC &rsqb;_) ._. 
</s>
<s>
They_PPHS2 use_VV0 a_AT1 learning_NN1 set_NN1 of_IO pre-categorized_JJ requirements_NN2 to_TO train_VVI a_AT1 classifier_NN1 ._. 
</s>
<s>
Future_JJ research_NN1 should_VM test_VVI the_AT model_NN1 in_II expanded_JJ populations_NN2 including_VVG in_II international_JJ contexts_NN2 ._. 
</s>
<s>
In_II this_DD1 section_NN1 ,_, we_PPIS2 discuss_VV0 each_DD1 component_NN1 in_II more_DAR detail_NN1 ._. 
</s>
<s>
One_MC1 of_IO the_AT implicit_JJ code_NN1 lists_NN2 is_VBZ the_AT age_NN1 categories_NN2 according_II21 to_II22 SDMX_NP1 ._. 
</s>
<s>
This_DD1 could_VM lead_VVI to_II a_AT1 higher_JJR probability_NN1 of_IO "_" selling_NN1 of_IO trade_NN1 secrets_NN2 "_" of_IO a_AT1 company_NN1 ,_, which_DDQ could_VM increase_VVI its_APPGE "_" competitive_JJ disadvantage_NN1 "_" and_CC therefore_RR reduce_VV0 the_AT company_NN1 '_NULL s_ZZ1 ability_NN1 to_TO sell_VVI or_CC market_VVI its_APPGE products_NN2 and_CC services_NN2 ._. 
</s>
<s>
So_RR ,_, not_XX only_RR are_VBR these_DD2 gender-neutral_JJ phrase_NN1 the_AT most_RRT searched_VVN for_IF (_( i.e._REX ,_, more_DAR impressions_NN2 )_) ,_, they_PPHS2 are_VBR the_AT most_RRT clicked_VVN on_II (_( i.e._REX ,_, have_VH0 the_AT most_DAT consumer_NN1 interest_NN1 )_) ._. 
</s>
<s>
Some_DD mutants_NN2 will_VM also_RR lead_VVI to_TO run_VVI time_NNT1 exceptions_NN2 (_( we_PPIS2 are_VBR ,_, after_II all_DB ,_, trying_VVG to_TO cause_VVI failures_NN2 !_! )_) ,_, which_DDQ in_II Unix_NN1 C_ZZ1 programs_NN2 usually_RR cause_VV0 the_AT process_NN1 to_TO terminate_VVI ._. 
</s>
<s>
In_RR21 addition_RR22 ,_, Section6describes_FO a_AT1 polynomial_NN1 kernel_NN1 for_IF the_AT consistency_NN1 problem_NN1 for_IF parameterized_VVD by_II the_AT number_NN1 of_IO holes_NN2 in_II the_AT variable_JJ domains_NN2 ,_, and_CC an_AT1 FPT_NP1 algorithm_NN1 that_CST uses_VVZ this_DD1 kernel_NN1 as_II a_AT1 first_MD step_NN1 ._. 
</s>
<s>
Other_JJ wetting_NN1 regions_NN2 (_( 4,5_MC )_) and_CC the_AT non-warming_JJ region_NN1 (_( 7_MC )_) have_VH0 less_DAR risk_NN1 increase_NN1 that_CST the_AT national_JJ average_NN1 ._. 
</s>
<s>
VDG_NP1 gives_VVZ us_PPIO2 that_DD1 information_NN1 since_CS a_AT1 SCD-altered_JJ protein_NN1 has_VHZ a_AT1 greatest_JJT impact_NN1 on_II proteins_NN2 in_II its_APPGE cluster_NN1 ._. 
</s>
<s>
Digital_JJ Preservation_NN1 is_VBZ an_AT1 interdisciplinary_JJ field_NN1 where_CS long-term_JJ change_NN1 insights_NN2 are_VBR crucial_JJ ,_, involving_VVG experts_NN2 in_II arts_NN2 ,_, culture_NN1 ,_, history_NN1 and_CC literature_NN1 &lsqb;_( 28_MC &rsqb;_) ._. 
</s>
<s>
People_NN within_II organizations_NN2 ,_, despite_II having_VHG the_AT intention_NN1 to_TO comply_VVI with_IW information_NN1 security_NN1 policies_NN2 ,_, are_VBR still_RR considered_VVN to_TO be_VBI the_AT weakest_JJT link_NN1 in_II defense_NN1 against_II the_AT existing_JJ information_NN1 security_NN1 as_CSA their_APPGE actual_JJ security_NN1 behavior_NN1 may_VM differ_VVI from_II the_AT intended_JJ behavior_NN1 (_( Han_NP1 et_RA21 al._RA22 ,_, 2008_MC ,_, Guo_NP1 et_RA21 al._RA22 ,_, 2011_MC ,_, Capelli_NP1 et_RA21 al._RA22 ,_, 2006_MC and_CC Vroom_NP1 and_CC Solms_NP1 ,_, 2004_MC )_) ._. 
</s>
<s>
In_II this_DD1 experiment_NN1 ,_, the_AT SLBP_NP1 has_VHZ 25_MC codes_NN2 (_( 24_MC ULBP_NN1 codes_NN2 and_CC 1_MC1 dummy_JJ code_NN1 )_) since_CS we_PPIS2 take_VV0 the_AT symmetry_NN1 levels_NN2 3_MC and_CC 4_MC in_II Eq_NN1 ._. 
</s>
<s>
(_( 3_MC )_) ._. 
</s>
<s>
The_AT edges_NN2 connecting_VVG these_DD2 states_NN2 represent_VV0 the_AT transition_NN1 probabilities_NN2 between_II these_DD2 blocks_NN2 &lsqb;_( 18_MC &rsqb;_) ,_, &lsqb;_( 19_MC &rsqb;_) ._. 
</s>
<s>
The_AT algorithm_NN1 iterates_VVZ over_II the_AT triple_JJ pattern_NN1 iteration_NN1 scope_NN1 of_IO the_AT addition_NN1 and_CC deletion_NN1 trees_NN2 in_II a_AT1 sort-merge_VV0 join-like_JJ operation_NN1 ,_, and_CC only_RR emits_VVZ the_AT triples_NN2 that_CST have_VH0 a_AT1 different_JJ addition/deletion_NN1 flag_NN1 for_IF the_AT two_MC versions_NN2 ._. 
</s>
<s>
The_AT decision_NN1 path_NN1 for_IF the_AT second_MD instance_NN1 in_II Table_NN1 1_MC1 is_VBZ :_: 01_MC →_NULL 02_MC →_NULL leaf(0.0)_FO ._. 
</s>
<s>
In_II the_AT ensuing_JJ sections_NN2 ,_, we_PPIS2 describe_VV0 the_AT specifics_NN2 of_IO each_DD1 markup_NN1 type_NN1 ,_, as_II31 well_II32 as_II33 universal_JJ class-independent_JJ markup_NN1 practices_NN2 ._. 
</s>
<s>
The_AT implementation_NN1 of_IO this_DD1 browser_NN1 includes_VVZ a_AT1 mechanism_NN1 that_CST displays_VVZ to_II the_AT user_NN1 justification_NN1 of_IO why_RRQ a_AT1 Web_NN1 site_NN1 should_VM be_VBI trusted_VVN ._. 
</s>
<s>
All_DB these_DD2 issues_NN2 occurred_VVD with_IW common_JJ JavaScript_NN1 libraries_NN2 like_II jQuery_NN1 as_II the_AT cacheable_JJ resource_NN1 and_CC were_VBDR subsequently_RR deemed_VVN false_JJ positives_NN2 ._. 
</s>
<s>
It_PPH1 is_VBZ therefore_RR unnecessary_JJ to_TO use_VVI any_DD locking_JJ mechanism_NN1 to_TO protect_VVI access_NN1 to_II the_AT sub-plans_NN2 ._. 
</s>
<s>
From_II the_AT beginning_NN1 ,_, the_AT objective_NN1 was_VBDZ to_TO define_VVI a_AT1 Software_NN1 Product_NN1 Line_NN1 (_( FSFB-SPL_NP1 )_) ,_, capable_JJ of_IO deriving_VVG different_JJ tele-consultation_JJ products_NN2 ,_, sharing_VVG a_AT1 common_JJ architecture_NN1 and_CC reusing_VVG a_AT1 set_NN1 of_IO components_NN2 developed_VVN to_II this_DD1 end_NN1 ._. 
</s>
<s>
A_AT1 simplified_JJ version_NN1 of_IO the_AT PReM_NN1 architecture_NN1 is_VBZ illustrated_VVN in_II Fig._NN1 8_MC ._. 
</s>
<s>
In_RR21 addition_RR22 ,_, system_NN1 engineers_NN2 can_VM use_VVI the_AT formal_JJ definition_NN1 of_IO the_AT TCF_NP1 policies_NN2 as_II a_AT1 guideline_NN1 to_TO design_VVI and_CC implement_VVI a_AT1 covert-timing_JJ channel_NN1 free_JJ scheduler_NN1 or_CC scheduling_VVG algorithm_NN1 ._. 
</s>
<s>
Past_JJ studies_NN2 (_( e.g._REX &lsqb;_( Mabert_NP1 et_RA21 al._RA22 ,_, 2000_MC &rsqb;_) ,_, &lsqb;_( Ifinedo_NP1 ,_, 2007_MC &rsqb;_) and_CC &lsqb;_( McGinnis_NP1 and_CC Huang_NP1 ,_, 2007_MC &rsqb;_) )_) and_CC the_AT commercial_JJ press_NN1 (_( &lsqb;_( Stedman_NP1 ,_, 1999_MC &rsqb;_) and_CC &lsqb;_( Songini_NP1 ,_, 2000_MC &rsqb;_) )_) suggest_VV0 that_CST many_DA2 organizations_NN2 are_VBR dissatisfied_JJ with_IW benefits_NN2 obtained_VVN from_II their_APPGE Enterprise_NN1 Systems_NN2 investments_NN2 ._. 
</s>
<s>
To_TO address_VVI this_DD1 problem_NN1 ,_, two_MC pre-replacement_JJ imputation_NN1 techniques_NN2 were_VBDR explored_VVN (_( Young_NP1 et@al._FO ,_, 2010_MC )_) ,_, which_DDQ included_VVD multiple_JJ imputation_NN1 (_( Schafer_NP1 ,_, 1997_MC )_) and_CC expectation_NN1 maximization_NN1 (_( Dempster_NP1 et@al._FO ,_, 1977_MC )_) ._. 
</s>
<s>
Zayed_JJ University_NN1 does_VDZ not_XX have_VHI an_AT1 IS_VBZ policy_NN1 ,_, an_AT1 IS_VBZ security_NN1 policy_NN1 ,_, and_CC made_VVD no_AT effort_NN1 to_TO educate_VVI its_APPGE users_NN2 about_II the_AT newly_RR issued_VVN laws_NN2 ._. 
</s>
<s>
This_DD1 enhances_VVZ the_AT external_JJ validity_NN1 of_IO the_AT research_NN1 findings_NN2 ._. 
</s>
<s>
The_AT paper_NN1 is_VBZ organized_VVN as_CSA follows_VVZ :_: the_AT next_MD section_NN1 starts_VVZ with_IW a_AT1 brief_JJ review_NN1 of_IO challenges_NN2 for_IF evaluating_JJ IAMs_NP2 ,_, most_DAT of_IO which_DDQ are_VBR connected_VVN with_IW the_AT properties_NN2 of_IO open_JJ systems_NN2 whose_DDQGE future_JJ behavior_NN1 is_VBZ fundamentally_RR unknown_JJ ._. 
</s>
<s>
In_II this_DD1 work_NN1 we_PPIS2 empirically_RR defined_VVN that_CST Wk=1_FO ,_, which_DDQ indicates_VVZ that_CST each_DD1 training_NN1 sequence_NN1 has_VHZ the_AT same_DA weight_NN1 ._. 
</s>
<s>
We_PPIS2 then_RT added_VVD a_AT1 Leave/Remain_VV0 classifier_NN1 ,_, described_VVN in_II Section_NN1 6.2_MC ,_, which_DDQ helped_VVD us_PPIO2 to_TO identify_VVI a_AT1 reliable_JJ sample_NN1 of_IO tweets_NN2 with_IW unambiguous_JJ stance_NN1 ._. 
</s>
<s>
As_II a_AT1 result_NN1 ,_, aproxynode_NN1 can_VM only_RR be_VBI pruned_VVN from_II the_AT tree_NN1 when_CS the_AT status_NN1 of_IO the_AT node_NN1 itself_PPX1 and_CC all_DB of_IO the_AT original_JJ nodes_NN2 referring_VVG to_II it_PPH1 change_VV0 tobelong_NN1 ._. 
</s>
<s>
The_AT superiority_NN1 of_IO ICRGMM_NP1 is_VBZ evaluated_VVN based_VVN on_II texture_NN1 clustering_NN1 and_CC Graph_NN1 Cuts_NN2 based_VVN texture_NN1 segmentation_NN1 using_VVG a_AT1 large_JJ number_NN1 of_IO synthesis_NN1 texture_NN1 images_NN2 and_CC real_JJ natural_JJ scene_NN1 textured_JJ images_NN2 ._. 
</s>
<s>
Instead_II21 of_II22 relying_VVG on_II the_AT volume_NN1 as_II the_AT main_JJ metric_JJ this_DD1 method_NN1 assumes_VVZ that_CST anomalies_NN2 will_VM disturb_VVI the_AT distribution_NN1 of_IO traffic_NN1 features_NN2 in_II a_AT1 detectable_JJ manner_NN1 ._. 
</s>
<s>
The_AT auctioneer_NN1 contacts_NN2 the_AT selected_JJ host_NN1 with_IW an_AT1 MREQ_JJ message_NN1 ,_, which_DDQ contains_VVZ the_AT request_NN1 for_IF a_AT1 surrogate_NN1 as_CSA discussed_VVN in_II Section_NN1 3.2_MC ._. 
</s>
<s>
Other_JJ studies_NN2 (_( Florêncio_NP1 and_CC Herley_NP1 ,_, 2007_MC )_) report_VV0 even_RR more_RGR extreme_JJ figures_NN2 where_CS in_II a_AT1 study_NN1 of_IO half_DB a_AT1 million_NNO Internet_NN1 users_NN2 each_DD1 was_VBDZ found_VVN on_II average_JJ to_TO have_VHI 25_MC passwords_NN2 ,_, reusing_VVG each_DD1 for_IF an_AT1 average_NN1 of_IO 6.5_MC different_JJ sites_NN2 ._. 
</s>
<s>
On_II this_DD1 point_NN1 the_AT results_NN2 are_VBR quite_RG clear_JJ ._. 
</s>
<s>
FR7_FO :_: the_AT tool_NN1 shall_VM allow_VVI users_NN2 to_TO refine_VVI their_APPGE assigned_JJ obligations_NN2 by_II creating_VVG new_JJ obligations_NN2 and_CC assigning_VVG those_DD2 obligations_NN2 to_II themselves_PPX2 or_CC other_JJ actors_NN2 ._. 
</s>
<s>
He_PPHS1 observes_VVZ four_MC common_JJ anti-patterns_NN2 ,_, which_DDQ are_VBR behaviors_NN2 that_CST most_DAT agree_VV0 are_VBR suboptimal_JJ ,_, but_CCB are_VBR still_RR the_AT norm_NN1 in_II today_RT '_NULL s_ZZ1 cyber_NN1 reality_NN1 ._. 
</s>
<s>
In_II the_AT definition_NN1 ,_, two_MC main_JJ items_NN2 are_VBR specified_VVN ._. 
</s>
<s>
In_RR21 general_RR22 ,_, such_DA an_AT1 approach_NN1 does_VDZ not_XX require_VVI knowledge_NN1 of_IO applications_NN2 running_VVG on_II web_NN1 servers_NN2 or_CC any_DD significant_JJ pre-processing_NN1 of_IO incoming_JJ data_NN ._. 
</s>
<s>
A_AT1 full_JJ update_NN1 ,_, at_II each_DD1 optimization_NN1 iteration_NN1 ,_, using_VVG forward_RL and/or_CC central_JJ finite_JJ differences_NN2 ._. 
</s>
<s>
As_II a_AT1 consequence_NN1 of_IO improved_JJ traffic_NN1 flow_NN1 ,_, the_AT scheme_NN1 offsets_VVZ the_AT emissions_NN2 from_II the_AT original_JJ 2003_MC traffic_NN1 flows_VVZ by_II around_RG 76_MC kilo-tonnes_NNU2 of_IO CO2_FO equivalency_NN1 (_( 4%_NNU reduction_NN1 )_) ._. 
</s>
<s>
The_AT index_NN1 embodies_VVZ the_AT intuitive_JJ notion_NN1 that_CST when_CS temperature_NN1 is_VBZ high_JJ and_CC relative_JJ humidity_NN1 is_VBZ low_JJ ,_, fuel_NN1 moisture_NN1 content_NN1 is_VBZ expected_VVN to_TO be_VBI relatively_RR low_JJ ,_, while_CS if_CS temperature_NN1 is_VBZ low_JJ and_CC relative_JJ humidity_NN1 is_VBZ high_JJ ,_, fuel_NN1 moisture_NN1 content_NN1 is_VBZ expected_VVN to_TO be_VBI relatively_RR high_JJ ._. 
</s>
<s>
In_RR21 addition_RR22 ,_, a_AT1 normalized_JJ ranking_NN1 method_NN1 was_VBDZ used_VVN to_TO differentiate_VVI the_AT concept_NN1 terms_NN2 and_CC general_JJ terms_NN2 ._. 
</s>
<s>
Computationally_RR ,_, we_PPIS2 are_VBR not_XX aware_JJ of_IO any_DD algorithm_NN1 that_CST can_VM perform_VVI such_DA an_AT1 exhaustive_JJ search_NN1 with_IW such_DA efficiency_NN1 ._. 
</s>
<s>
This_DD1 reduces_VVZ the_AT variation_NN1 between_II the_AT four_MC subgroups_NN2 ,_, but_CCB increases_VVZ each_DD1 subgroup_NN1 '_NULL s_ZZ1 internal_JJ variation_NN1 ._. 
</s>
<s>
The_AT investigated_JJ period_NN1 of_IO 15_MC years_NNT2 might_VM be_VBI too_RG short_JJ for_IF observing_VVG climate_NN1 change_NN1 induced_VVD increment_NN1 trends_NN2 ._. 
</s>
<s>
However_RR ,_, results_NN2 also_RR show_VV0 that_DD1 surface_NN1 runoff_NN1 for_IF models_NN2 with_IW non-calibrated_JJ storage_NN1 heights_NN2 is_VBZ generated_VVN too_RG early_RR in_II the_AT simulations_NN2 (_( Fig._NN1 ?_? 5_MC )_) (_( day_NNT1 19_MC and_CC 23_MC )_) compared_VVN to_II the_AT micro-topography_NN1 model_NN1 (_( day_NNT1 45_MC )_) ._. 
</s>
<s>
This_DD1 can_VM be_VBI a_AT1 common_JJ occurrence_NN1 in_II wireless_NN1 access_NN1 settings_NN2 and_CC at_II the_AT typically_RR constrained_VVN interfaces_NN2 linking_VVG ISP_NN1 networks_NN2 and_CC enterprise_NN1 networks_NN2 (_( peering_VVG links_NN2 )_) ._. 
</s>
<s>
The_AT BFI_NN1 ,_, ranging_VVG from_II 0_MC to_II 1.0_MC (_( Gustard_NP1 et_NP1 ?_? al._NNU ,_, 1992_MC )_) and_CC derived_VVD using_VVG hydrograph_NN1 analysis_NN1 of_IO long-term_JJ gauged_VVD daily_RR mean_JJ flows_NN2 ,_, is_VBZ a_AT1 measure_NN1 of_IO the_AT proportion_NN1 of_IO the_AT river_NN1 runoff_NN1 that_CST originates_VVZ from_II stored_JJ sources_NN2 ._. 
</s>
<s>
Trustguide_NN1 was_VBDZ a_AT1 collaborative_JJ project_NN1 between_II BT_NP1 Group_NN1 and_CC HP_NN1 Labs_NN2 ,_, aiming_VVG to_TO better_RRR understand_VVI public_JJ attitudes_NN2 toward_II online_JJ security_NN1 in_BCL21 order_BCL22 to_TO develop_VVI more_RGR effective_JJ ICT-based_JJ services_NN2 (_( Lacohee_NP1 et@al._FO ,_, 2006_MC )_) ._. 
</s>
<s>
Over_II the_AT test_NN1 images_NN2 the_AT EJS_NP1 jet_NN1 positions_NN2 are_VBR shown_VVN as_II a_AT1 bright_JJ dot_NN1 ._. 
</s>
<s>
Analog_VV0 signs_NN2 are_VBR projected_VVN into_II a_AT1 categorical_JJ signs_NN2 ._. 
</s>
<s>
Ambiguous_JJ and_CC irrelevant_JJ questions_NN2 are_VBR typically_RR discarded_VVN ,_, as_CSA they_PPHS2 are_VBR difficult_JJ to_TO fix_VVI ._. 
</s>
<s>
A_AT1 two-point_JJ rack_NN1 enables_VVZ to_TO extract_VVI the_AT boundary_NN1 as_CSA shown_VVN on_II the_AT top_NN1 ,_, which_DDQ is_VBZ hardly_RR possible_JJ to_TO deal_VVI with_IW using_VVG the_AT traditional_JJ dynamic_JJ programming_NN1 ._. 
</s>
<s>
Let_VV0 Pd_NP1 denote_VVI the_AT pseudo-diameter_NN1 of_IO a_AT1 source_NN1 ._. 
</s>
<s>
With_IW join=fail=_FO ,_, a_AT1 stable_JJ number_NN1 ofT-nodes_NN2 over_II time_NNT1 indicates_VVZ that_CST our_APPGE protocols_NN2 were_VBDR effective_JJ and_CC stable_JJ ._. 
</s>
<s>
FTP_NP1 ,_, Internet_NP1 ,_, or_CC dedicated_JJ private_JJ communication_NN1 networks_NN2 ._. 
</s>
<s>
Once_CS the_AT policy_NN1 system_NN1 has_VHZ decided_VVN which_DDQ actions_NN2 apply_VV0 (_( a_AT1 possibly_RR empty_JJ list_NN1 )_) ,_, these_DD2 are_VBR sent_VVN to_II the_AT system_NN1 interface_NN1 for_IF execution_NN1 ._. 
</s>
<s>
The_AT derivation_NN1 of_IO the_AT CUB_NN1 resembles_VVZ that_DD1 of_IO the_AT Naive_JJ bound_NN1 ._. 
</s>
<s>
Unfortunately_RR ,_, the_AT story_NN1 as_CSA relayed_VVN above_RL would_VM simply_RR give_VVI a_AT1 degenerate_JJ (_( complete_JJ )_) BN_NP1 ._. 
</s>
<s>
It_PPH1 seems_VVZ reasonable_JJ to_TO define_VVI the_AT best_JJT choice_NN1 of_IO memory_NN1 length_NN1 as_CSA one_PN1 which_DDQ has_VHZ a_AT1 low_JJ volatility_NN1 and_CC produces_VVZ a_AT1 mean_JJ and_CC maximum_JJ emission_NN1 value_NN1 that_CST is_VBZ close_JJ to_II the_AT monthly_JJ limit_NN1 ._. 
</s>
<s>
Since_CS an_AT1 ontology_NN1 is_VBZ a_AT1 shared_JJ conceptualization_NN1 ,_, annotations_NN2 made_VVD using_VVG its_APPGE concepts_NN2 reduce_VV0 the_AT ambiguousness_NN1 of_IO syntactical_JJ service_NN1 contracts_NN2 ._. 
</s>
<s>
The_AT most_RGT costly_JJ solution_NN1 is_VBZ the_AT simple_JJ attestation_NN1 ;_; this_DD1 stands_VVZ to_TO reason_VVI as_II every_AT1 (_( uncached_JJ )_) UPDATE_NN1 leads_VVZ to_II a_AT1 signature_NN1 validation_NN1 ._. 
</s>
<s>
Finally_RR ,_, '_NULL GnuTLS_VVZ via_II AOP_NP1 '_NULL shows_VVZ the_AT performance_NN1 of_IO using_VVG AspectC++_FO to_TO inject_VVI code_NN1 ,_, which_DDQ is_VBZ equivalent_JJ to_II '_NULL GnuTLS_NP1 Hash_NP1 Table_NN1 '_NULL ._. 
</s>
<s>
This_DD1 section_NN1 describes_VVZ the_AT architecture_NN1 and_CC main_JJ components_NN2 of_IO the_AT policy_NN1 system_NN1 ._. 
</s>
<s>
All_DB the_AT alarm_NN1 bells_NN2 were_VBDR ringing_VVG ._. 
</s>
<s>
From_II the_AT information_NN1 we_PPIS2 collected_VVD ,_, we_PPIS2 noticed_VVD that_CST the_AT generation_NN1 of_IO the_AT tables_NN2 was_VBDZ significantly_RR longer_RRR than_CSN any_DD other_JJ process_NN1 of_IO RainbowCrack_NN1 ._. 
</s>
<s>
However_RGQV transactive_JJ memory_NN1 alone_RR also_RR exerts_VVZ an_AT1 influence_NN1 of_IO effectiveness_NN1 ._. 
</s>
<s>
Pseudo-tags_NN2 are_VBR learned_VVN by_II aggregating_VVG tags_NN2 from_II similar_JJ tracks_NN2 ;_; i.e._REX tracks_NN2 whose_DDQGE audio_JJ content_NN1 is_VBZ similar_JJ according_II21 to_II22 the_AT MFS_NP1 music_NN1 inspired_JJ texture_NN1 representation_NN1 ._. 
</s>
<s>
We_PPIS2 used_VVD the_AT ORL_NP1 (_( Olivetti_NP1 Research_NN1 Laboratory_NN1 )_) faces_VVZ database_NN1 &lsqb;_( 54_MC &rsqb;_) for_IF verification_NN1 ._. 
</s>
<s>
We_PPIS2 presented_VVD a_AT1 scalable_JJ testing_NN1 and_CC evaluation_NN1 platform_NN1 ._. 
</s>
<s>
The_AT model_NN1 will_VM also_RR be_VBI used_VVN to_TO perform_VVI a_AT1 sensitivity_NN1 analysis_NN1 of_IO the_AT results_NN2 to_II changes_NN2 in_II key_JJ variables_NN2 ._. 
</s>
<s>
A_AT1 comparison_NN1 of_IO the_AT proposed_JJ model_NN1 with_IW previous_JJ works_NN shows_VVZ that_CST the_AT Markov_NP1 Chain_NN1 Model_NN1 outperforms_VVZ other_JJ approaches_NN2 that_CST use_VV0 the_AT Rayleigh_NP1 model_NN1 ?_? &lsqb;_( 7_MC &rsqb;_) ?_? to_TO predict_VVI node_NN1 distribution_NN1 ,_, being_VBG able_JK to_TO increase_VVI stability_NN1 in_II the_AT construction_NN1 of_IO clustered_JJ networks_NN2 ?_? &lsqb;_( 8_MC &rsqb;_) ._. 
</s>
<s>
Because_CS they_PPHS2 are_VBR written_VVN in_II the_AT same_DA formalism_NN1 and_CC are_VBR using_VVG the_AT same_DA (_( daily_RR )_) time_VV0 step_NN1 ,_, they_PPHS2 could_VM be_VBI easily_RR joined_VVN together_RL in_II a_AT1 single_JJ library_NN1 ._. 
</s>
<s>
Broadly_RR ,_, consultancies_NN2 that_CST advance_VV0 innovating_JJ with_IW IT_NN1 may_VM serve_VVI the_AT achievement_NN1 of_IO competitive_JJ parity_NN1 among_II firms_NN2 ,_, more_RRR than_CSN they_PPHS2 do_VD0 the_AT achievement_NN1 of_IO competitive_JJ advantage_NN1 ._. 
</s>
<s>
The_AT transitivity_NN1 of_IO the_AT subsumption_NN1 relation_NN1 can_VM also_RR be_VBI used_VVN to_TO remove_VVI obvious_JJ non-subsumptions_NN2 from_II P._NP1 All_DB that_DD1 is_VBZ necessary_JJ is_VBZ to_TO sum_VVI together_RL all_DB the_AT simulations_NN2 results_VVZ to_TO obtain_VVI the_AT frequency_NN1 of_IO occurrence_NN1 as_II a_AT1 result_NN1 of_IO urban_JJ development_NN1 in_II each_DD1 cell_NN1 ._. 
</s>
<s>
It_PPH1 turns_VVZ out_RP that_DD1 kernel_NN1 performance_NN1 differs_VVZ per_II dataset_NN1 ._. 
</s>
<s>
I_PPIS1 begin_VV0 by_II randomly_RR setting_VVG the_AT values_NN2 of_IO q=0.3_FO ,_, =0.1_FO ,_, and_CC =0.01_FO and_CC experiment_NN1 with_IW the_AT best_JJT value_NN1 for_IF γ_NULL in_II set_NN1 0.5_MC ,_, 1_MC1 ,_, 2_MC ._. 
</s>
<s>
The_AT application_NN1 of_IO the_AT HEAM_NN1 approach_NN1 is_VBZ demonstrated_VVN here_RL with_IW the_AT help_NN1 of_IO an_AT1 agile_JJ EA_NP1 modelling_NN1 case_NN1 study_NN1 example_NN1 involving_VVG ArchiMate_NP1 ,_, BPMN_NP1 and_CC UML_NP1 ._. 
</s>
<s>
MADE_VVN was_VBDZ accessible_JJ but_CCB people_NN have_VH0 problems_NN2 with_IW it_PPH1 ,_, we_PPIS2 shouldn_VV0 '_NULL t_ZZ1 underestimate_VV0 the_AT challenge_NN1 of_IO making_VVG sense_NN1 of_IO figures_NN2 even_CS21 when_CS22 the_AT maths_NN1 is_VBZ stripped_VVN out_RP ._. 
</s>
<s>
A_AT1 key_JJ point_NN1 in_II designing_VVG resource-oriented_JJ interfaces_NN2 is_VBZ the_AT identification_NN1 and_CC description_NN1 of_IO the_AT state_NN1 transitions_NN2 to_TO support_VVI linkages_NN2 between_II resources_NN2 ._. 
</s>
<s>
In_BCL21 order_BCL22 to_TO test_VVI and_CC validate_VVI the_AT potential_JJ benefits_NN2 of_IO ViGOR_NN1 in_II assisting_VVG with_IW video_NN1 search_NN1 ,_, we_PPIS2 conducted_VVD two_MC user_NN1 studies_NN2 ,_, in_II which_DDQ we_PPIS2 tested_VVD two_MC systems_NN2 ._. 
</s>
<s>
Similarly_RR ,_, network_NN1 resources_NN2 may_VM be_VBI virtualized_VVN into_II GENI_JJ slices_NN2 using_VVG two_MC different_JJ paradigms_NN2 ._. 
</s>
<s>
As_CS31 far_CS32 as_CS33 unweighted_JJ hypergraphs_NN2 are_VBR concerned_JJ ,_, the_AT methods_NN2 reported_VVN in_II the_AT literature_NN1 mainly_RR focus_VV0 on_II using_VVG tensors_NN2 to_TO represent_VVI hypergraphs_NN2 &lsqb;_( 31_MC &rsqb;_) ,_, &lsqb;_( 32_MC &rsqb;_) ,_, &lsqb;_( 38_MC &rsqb;_) ._. 
</s>
<s>
For_IF each_DD1 index_NN1 ,_, its_APPGE maximum_JJ BDC_NP1 –_- either_RR including_VVG or_CC excluding_VVG the_AT category_NN1 to_II which_DDQ the_AT index_NN1 belongs_VVZ –_- has_VHZ been_VBN summarized_VVN to_TO report_VVI the_AT highest_RRT detected_VVN dependence_NN1 between_II that_DD1 index_NN1 and_CC the_AT others_NN2 ._. 
</s>
<s>
Several_DA2 works_NN such_DA sending_VVG multilevel_NN1 redundant_JJ information_NN1 which_DDQ will_VM eventually_RR increase_VVI the_AT burden_NN1 on_II the_AT network_NN1 ._. 
</s>
<s>
We_PPIS2 may_VM calculate_VVI these_DD2 values_NN2 effectively_RR ,_, using_VVG hash_NN1 table_NN1 retrieval_NN1 as_II a_AT1 surrogate_NN1 for_IF distinguishing_VVG and_CC counting_VVG common_JJ and_CC distinct_JJ features_NN2 within_II the_AT setsTTi_NN2 ,_, T_ZZ1 Ti_NP1 ,_, and_CC Ti_NP1 T_ZZ1 respectively_RR ._. 
</s>
<s>
Considering_CS the_AT practical_JJ importance_NN1 of_IO the_AT problem_NN1 in_II recent_JJ years_NNT2 ,_, heuristic_JJ implementations_NN2 exist_VV0 ._. 
</s>
<s>
Finally_RR ,_, these_DD2 results_NN2 are_VBR compared_VVN to_II the_AT groundtruth_NN1 sequence_NN1 segmentations_NN2 ,_, and_CC the_AT error_NN1 rates_NN2 of_IO the_AT considered_JJ methods_NN2 are_VBR calculated_VVN ._. 
</s>
<s>
In_II astronomy_NN1 and_CC astrophysics_NN2 ,_, the_AT combination_NN1 of_IO software_NN1 models_NN2 and_CC advances_VVZ in_II digital_JJ imaging_NN1 systems_NN2 have_VH0 combined_VVN to_TO allow_VVI scientists_NN2 to_TO discover_VVI new_JJ solar_JJ systems_NN2 that_CST are_VBR too_RG faint_JJ for_IF human_JJ detection_NN1 ._. 
</s>
<s>
A_AT1 planar_JJ model_NN1 (_( surface_NN1 slope_NN1 0.03_MC ?_? m/m_ZZ1 )_) without_IW rill/depression_NN1 storage_NN1 height_NN1 variations_NN2 is_VBZ set_VVN up_RP using_VVG the_AT same_DA grid_NN1 resolution_NN1 as_CSA the_AT micro-topography_NN1 model_NN1 to_TO simulate_VVI reference_NN1 conditions_NN2 without_IW any_DD representation_NN1 of_IO micro-topography_NN1 ._. 
</s>
<s>
Note_VV0 that_CST the_AT parameters_NN2 in_II the_AT decoder_NN1 and_CC the_AT encoder_NN1 are_VBR not_XX sharing_VVG ._. 
</s>
<s>
By_II extending_VVG from_II the_AT ?_? hasAssociation_NN1 ?_? property_NN1 ,_, we_PPIS2 are_VBR asserting_VVG that_CST there_EX is_VBZ an_AT1 association_NN1 from_II one_MC1 concept_NN1 to_II another_DD1 but_CCB that_DD1 association_NN1 can_VM not_XX be_VBI described_VVN using_VVG any_DD of_IO the_AT standard_JJ OT_NN1 properties_NN2 ._. 
</s>
<s>
The_AT exact_JJ purpose_NN1 and_CC functionality_NN1 of_IO each_DD1 module_NN1 within_II the_AT framework_NN1 is_VBZ explained_VVN below_RL ._. 
</s>
<s>
Forrester_NP1 survey_NN1 data_NN consistently_RR show_VV0 that_DD1 investment_NN1 in_II ES_NP1 and_CC enterprise_NN1 applications_NN2 in_RR21 general_RR22 remains_VVZ the_AT top_JJ IT_NN1 spending_NN1 priority_NN1 ,_, with_IW the_AT ES_NP1 market_NN1 estimated_VVN at_II $38_NNU billion_NNO and_CC predicted_VVN to_TO grow_VVI at_II a_AT1 steady_JJ rate_NN1 of_IO 6.9%_FO reaching_VVG $50_NNU billion_NNO by_II 2012_MC (_( Wang_NP1 and_CC Hamerman_NP1 ,_, 2008_MC )_) ._. 
</s>
<s>
The_AT second_MD contribution_NN1 of_IO this_DD1 paper_NN1 shows_VVZ the_AT good_JJ relation_NN1 between_II PS_NN1 and_CC PG_NP1 in_II obtaining_VVG hybrid_JJ algorithms_NN2 that_CST allow_VV0 us_PPIO2 to_TO find_VVI very_RG promising_JJ solutions_NN2 ._. 
</s>
<s>
First_MD the_AT function_NN1 begins_VVZ a_AT1 loop_NN1 that_CST repeats_VVZ for_IF the_AT number_NN1 of_IO partitions_NN2 (_( found_VVN in_II PartitionCount_NP1 )_) ._. 
</s>
<s>
Therefore_RR ,_, we_PPIS2 decided_VVD to_TO investigate_VVI five_MC national_JJ CMS_NP1 ,_, which_DDQ are_VBR deployed_VVN nationwide_RL on_II the_AT large_JJ scale_NN1 and_CC have_VH0 publicly_RR available_JJ specifications_NN2 ._. 
</s>
<s>
The_AT model_NN1 showed_VVD superiority_NN1 in_II those_DD2 aspects_NN2 for_IF different_JJ events_NN2 scenarios_NN2 ,_, especially_RR in_II its_APPGE detection_NN1 ability_NN1 ._. 
</s>
<s>
Other_JJ standard_NN1 R_ZZ1 methods_NN2 are_VBR provided_VVN for_IF convenience_NN1 to_TO work_VVI with_IW hydromad_NN1 objects_NN2 :_: for_REX21 example_REX22 one_PN1 can_VM extract_VVI parameter_NN1 values_NN2 with_IW coef()_NN1 ,_, and_CC model_NN1 results_NN2 with_IW fitted()_JJ ,_, observed()_VVD ,_, or_CC residuals()_NN2 ._. 
</s>
<s>
Denote_VV0 by_II n_ZZ1 the_AT number_NN1 of_IO triples_NN2 present_VV0 in_II the_AT RDF_NP1 graph_NN1 ._. 
</s>
<s>
The_AT authors_NN2 the_AT bandwidth_NN1 allocation_NN1 problem_NN1 as_II an_AT1 integer_NN1 linear_JJ program_NN1 and_CC a_AT1 10-approximation_JJ algorithm_NN1 is_VBZ introduced_VVN by_II solving_VVG the_AT linear_JJ program_NN1 relaxation_NN1 and_CC rounding_VVG the_AT fractional_JJ solution_NN1 ._. 
</s>
<s>
Processes_NN2 provide_VV0 quantitative_JJ descriptions_NN2 of_IO the_AT relations_NN2 they_PPHS2 represent_VV0 as_CSA one_MC1 or_CC more_DAR equations_NN2 ._. 
</s>
<s>
Experimental_JJ fires_NN2 in_II WA_NP1 were_VBDR ignited_VVN along_II the_AT full_JJ length_NN1 of_IO the_AT plot_NN1 (_( 200_MC ?_? m_ZZ1 )_) using_VVG a_AT1 vehicle_NN1 mounted_JJ flamethrower_NN1 ._. 
</s>
<s>
The_AT participants_NN2 were_VBDR mostly_RR postgraduate_NN1 students_NN2 and_CC researchers_NN2 at_II our_APPGE university_NN1 ._. 
</s>
<s>
As_II a_AT1 result_NN1 ,_, the_AT number_NN1 of_IO vehicles_NN2 and_CC the_AT average_JJ speed_NN1 per_II road_NN1 segment_NN1 and_CC per_II time_NNT1 of_IO day_NNT1 were_VBDR provided_VVN ,_, which_DDQ are_VBR needed_VVN for_IF the_AT emission_NN1 modeling_NN1 ._. 
</s>
<s>
The_AT conceptualization_NN1 portrays_VVZ both_RR an_AT1 optimistic_JJ ,_, as_II31 well_II32 as_II33 a_AT1 darker_JJR side_NN1 ,_, of_IO globalization_NN1 ._. 
</s>
<s>
From_II hereon_NN1 we_PPIS2 may_VM write_VVI A_ZZ1 →_NULL S_ZZ1 B_ZZ1 to_TO denote_VVI that_CST A_ZZ1 defeatsS_VV0 B_ZZ1 and_CC A_ZZ1 S_ZZ1 B_ZZ1 to_TO denote_VVI that_CST A_ZZ1 does_VDZ not_XX defeatS_NN2 B._NP1 We_PPIS2 use_VV0 these_DD2 codes_NN2 throughout_II the_AT remainder_NN1 of_IO this_DD1 paper_NN1 to_TO refer_VVI to_II individual_JJ papers_NN2 ._. 
</s>
<s>
There_EX are_VBR different_JJ types_NN2 of_IO moments_NN2 and_CC they_PPHS2 can_VM be_VBI classified_VVN as_CSA non-orthogonal_JJ and_CC orthogonal_JJ moments_NN2 depending_II21 on_II22 the_AT basis_NN1 function_NN1 used_VVD ._. 
</s>
<s>
However_RR ,_, similarly_RR to_II the_AT other_JJ existing_JJ proposals_NN2 ,_, the_AT scoring_NN1 mechanism_NN1 is_VBZ not_XX supported_VVN by_II an_AT1 automatic_JJ tool_NN1 ._. 
</s>
<s>
Many_DA2 employees_NN2 did_VDD not_XX have_VHI offices_NN2 as_CSA they_PPHS2 worked_VVD mostly_RR from_II home_NN1 but_CCB when_CS they_PPHS2 needed_VVD to_TO come_VVI to_II the_AT office_NN1 they_PPHS2 could_VM book_NN1 one_MC1 of_IO the_AT "_" flexible_JJ office_NN1 "_" spaces_NN2 ._. 
</s>
<s>
Thus_RR ,_, one_PN1 can_VM draw_VVI a_AT1 realistic_JJ picture_NN1 of_IO how_RRQ the_AT network_NN1 is_VBZ behaving_VVG ._. 
</s>
<s>
Fig._NN1 4_MC shows_NN2 how_RRQ the_AT level_NN1 to_II which_DDQ users_NN2 needed_VVN to_TO scroll_VVI is_VBZ distributed_VVN ._. 
</s>
<s>
CUDA_NP1 threads_NN2 may_VM access_VVI data_NN from_II different_JJ device_NN1 memory_NN1 spaces_NN2 during_II their_APPGE execution_NN1 :_: registers_NN2 ,_, local_JJ ,_, shared_VVD ,_, global_JJ ,_, constant_JJ ,_, and_CC texture_NN1 memory_NN1 ._. 
</s>
<s>
Each_DD1 document_NN1 was_VBDZ analyzed_VVN to_TO count_VVI all_DB citations_NN2 made_VVN to_II each_DD1 work_NN1 listed_VVN in_II the_AT references_NN2 section_NN1 ._. 
</s>
<s>
The_AT results_NN2 shown_VVN in_II Table_NN1 1_MC1 indicate_VV0 that_CST the_AT cost_NN1 of_IO provisioning_NN1 self-healing_NN1 using_VVG NPC_NP1 is_VBZ always_RR lower_JJR than_CSN that_DD1 using_VVG 1_MC1 +_FO 1_MC1 protection_NN1 ,_, and_CC the_AT saving_NN1 in_II the_AT protection_NN1 resources_NN2 can_VM reach_VVI up_RG21 to_RG22 30%_NNU ._. 
</s>
<s>
Let3SAT()_FO (_( where_RRQ is_VBZ an_AT1 arbitrary_JJ parameterization_NN1 )_) denote_VV0 the_AT problemSAT()restricted_JJ to_II 3CNF_FO formulas_NN2 ._. 
</s>
<s>
Only_JJ time_NNT1 when_RRQ efficiency_NN1 is_VBZ not_XX one_MC1 of_IO the_AT most_RGT frequently_RR used_JJ criteria_NN2 is_VBZ while_CS evaluating_VVG the_AT use_NN1 of_IO software_NN1 patterns_NN2 during_II software_NN1 construction_NN1 ._. 
</s>
<s>
Relevant_JJ features_NN2 of_IO the_AT target_NN1 concept_NN1 are_VBR highlighted_VVN as_CSA light_JJ blue_JJ to_TO help_VVI the_AT teachers_NN2 ._. 
</s>
<s>
The_AT paper_NN1 considers_VVZ how_RRQ value_NN1 creation_NN1 through_II comparison_NN1 websites_NN2 can_VM be_VBI explained_VVN through_II the_AT three_MC generic_JJ value_NN1 configurations_NN2 identified_VVN by_II Stabell_NP1 and_CC Fjeldstad_NP1 ;_; the_AT value_NN1 chain_NN1 ,_, the_AT value_NN1 shop_NN1 and_CC the_AT value_NN1 network_NN1 (_( 1998_MC )_) ._. 
</s>
<s>
The_AT Australian_JJ government_NN1 '_NULL s_ZZ1 study_NN1 of_IO cybercrime_NN1 in_II communities_NN2 led_VVN to_II a_AT1 series_NN of_IO strategically_RR focused_VVN recommendations_NN2 grouped_VVN in_II seven_MC major_JJ areas_NN2 (_( Commonwealth_NN1 of_IO Australia_NP1 ,_, 2010a_FO )_) ._. 
</s>
<s>
Previous_JJ results_NN2 &lsqb;_( 22_MC &rsqb;_) ,_, &lsqb;_( 23_MC &rsqb;_) ,_, &lsqb;_( 24_MC &rsqb;_) and_CC &lsqb;_( 25_MC &rsqb;_) suggest_VV0 that_CST there_EX is_VBZ a_AT1 correlation_NN1 between_II the_AT number_NN1 of_IO eigenPulse_NN1 attributes_NN2 and_CC performance_NN1 ._. 
</s>
<s>
The_AT lifecycle_NN1 of_IO resources_NN2 ,_, e.g._REX when_CS they_PPHS2 are_VBR created_VVN and_CC used_VVN ,_, how_RRQ they_PPHS2 are_VBR transformed_VVN and_CC versioned_VVN ,_, is_VBZ crucial_JJ to_II provenance_NN1 ,_, as_CSA expressed_VVN by_II the_AT following_JJ requirements_NN2 ._. 
</s>
<s>
The_AT interaction_NN1 summary_NN1 view_NN1 is_VBZ to_TO give_VVI the_AT dashboard_NN1 '_NULL s_ZZ1 users_NN2 an_AT1 overview_NN1 of_IO a_AT1 completed_JJ interaction_NN1 between_II a_AT1 consumer_NN1 and_CC a_AT1 practitioner_NN1 ._. 
</s>
<s>
In_RR21 addition_RR22 p=l_FO denotes_VVZ the_AT local_JJ process/domain_NN1 ._. 
</s>
<s>
Future_JJ growth_NN1 can_VM be_VBI expected_VVN in_II the_AT areas_NN2 of_IO serious_JJ games_NN2 and_CC transmedia_NN1 for_IF education_NN1 and_CC training_NN1 as_CSA momentum_NN1 gains_NN2 and_CC these_DD2 approaches_NN2 become_VV0 more_RGR mainstream_JJ ._. 
</s>
<s>
This_DD1 process_NN1 terminates_VVZ when_RRQ an_AT1 explanation_NN1 for_IF the_AT problem_NN1 in_II the_AT service_NN1 ticket_NN1 is_VBZ established_VVN ._. 
</s>
<s>
The_AT users_NN2 on_II the_AT ground_NN1 need_NN1 to_TO know_VVI ._. "_" 
</s>
<s>
It_PPH1 enables_VVZ integrated_JJ modelling_NN1 between_II third-party_JJ components_NN2 and_CC models_NN2 that_CST are_VBR implemented_VVN as_CSA local_JJ ,_, OpenMI-compliant_JJ components_NN2 (_( Gregersen_NP1 et_NP1 ?_? al._NNU ,_, 2007_MC )_) ._. 
</s>
<s>
Research_NN1 findings_NN2 drawn_VVN from_II works_NN examining_VVG such_DA sources_NN2 are_VBR particularly_RR valid_JJ if_CS the_AT data_NN represents_VVZ the_AT primary_JJ means_NN of_IO interaction_NN1 and_CC so_RR captures_VVZ team_NN1 processes_NN2 during_II software_NN1 development_NN1 &lsqb;_( 10_MC &rsqb;_) ._. 
</s>
<s>
During_II individual_JJ interviews_NN2 ,_, the_AT red_JJ team_NN1 members_NN2 became_VVD familiar_JJ with_IW terminology_NN1 and_CC concepts_NN2 ._. 
</s>
<s>
Their_APPGE cardinality_NN1 should_VM be_VBI sufficiently_RR small_JJ to_TO reduce_VVI both_DB2 the_AT storage_NN1 and_CC evaluation_NN1 time_NNT1 spent_VVN by_II an_AT1 NN_NP1 classifier_NN1 ._. 
</s>
<s>
The_AT web_NN1 browser_NN1 client_NN1 can_VM display_VVI web_NN1 pages_NN2 in_II HTML_NP1 ,_, execute_VV0 mobile_JJ code_NN1 such_II21 as_II22 Java_NP1 applets_NN2 or_CC web_NN1 scripts_NN2 ,_, and_CC capture_VV0 business-specific_JJ semantics_NN1 using_VVG XML_NP1 ._. 
</s>
<s>
The_AT user_NN1 should_VM have_VHI control_NN1 over_II how_RGQ much_DA1 of_IO this_DD1 data_NN can_VM be_VBI shared_VVN within_II the_AT space_NN1 ._. 
</s>
<s>
The_AT first_MD application_NN1 of_IO the_AT framework_NN1 was_VBDZ based_VVN on_II communes_NN2 ._. 
</s>
<s>
Our_APPGE decision_NN1 is_VBZ justified_VVN by_II a_AT1 factor_NN1 analysis_NN1 on_II the_AT items_NN2 ,_, which_DDQ found_VVD that_CST a_AT1 four-factor_JJ model_NN1 fitted_VVN the_AT data_NN significantly_RR better_RRR than_CSN a_AT1 one-factor_JJ solution_NN1 and_CC that_CST the_AT four_MC factors_NN2 corresponded_VVN to_II the_AT four_MC separate_JJ dimensions_NN2 of_IO XP_NP1 (_( this_DD1 is_VBZ reported_VVN in_II Michaelides_NP1 et_RA21 al_RA22 ._. 
</s>
<s>
&lsqb;_( 56_MC &rsqb;_) )_) ._. 
</s>
<s>
Nevertheless_RR ,_, their_APPGE wide_JJ range_NN1 of_IO applications_NN2 has_VHZ motivated_VVN large_JJ efforts_NN2 to_TO better_RRR understand_VVI these_DD2 systems_NN2 and_CC their_APPGE numerical_JJ approximations_NN2 ._. 
</s>
<s>
Such_DA communication_NN1 can_VM reduce_VVI rationalization_NN1 behaviors_NN2 and_CC ultimately_RR intentions_NN2 to_TO violate_VVI policy_NN1 by_II helping_VVG employees_NN2 realize_VVI that_CST rationalizations_NN2 are_VBR faulty_JJ and_CC that_CST the_AT associated_JJ behavior_NN1 is_VBZ not_XX "_" normal_JJ ._. "_" 
</s>
<s>
Motivation_NN1 :_: To_TO determine_VVI allowed_JJ actions_NN2 based_VVN on_II dynamically_RR changing_JJ properties_NN2 of_IO a_AT1 subject_NN1 or_CC object_VV0 instead_II21 of_II22 identities_NN2 ._. 
</s>
<s>
Over_II the_AT last_MD several_DA2 years_NNT2 different_JJ incremental_JJ algorithms_NN2 have_VH0 been_VBN also_RR proposed_VVN to_TO solve_VVI it_PPH1 (_( see_VV0 &lsqb;_( 1_MC1 &rsqb;_) ,_, &lsqb;_( 2_MC &rsqb;_) ,_, &lsqb;_( 3_MC &rsqb;_) ,_, &lsqb;_( 4_MC &rsqb;_) )_) ._. 
</s>
<s>
The_AT availability_NN1 of_IO standards_NN2 has_VHZ been_VBN a_AT1 critical_JJ factor_NN1 to_TO enable_VVI this_DD1 transition_NN1 to_II a_AT1 Web_NN1 of_IO data_NN ._. 
</s>
<s>
As_CSA it_PPH1 can_VM be_VBI seen_VVN ,_, both_DB2 Kernel_NP1 Estimators_NP1 performed_VVD extremely_RR well_RR in_II a_AT1 realistic_JJ implementation_NN1 in_II two_MC different_JJ scenarios_NN2 ._. 
</s>
<s>
A_AT1 linear_JJ order_NN1 on_II C_ZZ1 is_VBZ a_AT1 transitive_JJ ,_, antisymmetric_JJ ,_, and_CC total_JJ relation_NN1 on_II C._NP1 One_MC1 simple_JJ solution_NN1 is_VBZ to_TO store_VVI them_PPHO2 sorted_VVD by_II docID_NN1 ._. 
</s>
<s>
Agents_NN2 may_VM carry_VVI histories_NN2 of_IO past_JJ interactions_NN2 with_IW other_JJ agents_NN2 ._. 
</s>
<s>
The_AT results_NN2 of_IO this_DD1 study_NN1 should_VM not_XX be_VBI taken_VVN as_CSA generalized_JJ facts_NN2 ,_, but_CCB rather_RR the_AT interpretations_NN2 of_IO some_DD red_JJ team_NN1 members_NN2 '_NULL evaluations_NN2 of_IO CIS_NN2 systems_NN2 ._. 
</s>
<s>
CaDAnCE_NP1 demonstrated_VVD that_CST dependencies_NN2 among_II these_DD2 sub-applications_NN2 can_VM yield_VVI deployment-order_JJ priority_NN1 inversions_NN2 where_RRQ low-priority_JJ applications_NN2 may_VM complete_VVI their_APPGE deployments_NN2 ahead_II21 of_II22 a_AT1 mission-critical_JJ sub-application_NN1 ._. 
</s>
<s>
The_AT point_NN1 of_IO the_AT discussion_NN1 in_II this_DD1 section_NN1 is_VBZ that_CST a_AT1 particular_JJ pair_NN of_IO AE_FO and_CC CV_NN1 values_NN2 can_VM not_XX be_VBI considered_VVN definitive_JJ without_IW taking_VVG into_II account_NN1 how_RRQ representative_NN1 they_PPHS2 are_VBR within_II a_AT1 range_NN1 of_IO values_NN2 that_CST could_VM be_VBI produced_VVN for_IF the_AT same_DA game_NN1 under_II either_DD1 slightly_RR different_JJ conditions_NN2 (_( such_II21 as_II22 a_AT1 different_JJ analysis_NN1 depth_NN1 )_) or_CC even_RR the_AT same_DA conditions_NN2 (_( in_II a_AT1 multi-threaded_JJ environment_NN1 )_) ._. 
</s>
<s>
This_DD1 section_NN1 presents_VVZ the_AT hybridization_NN1 model_NN1 that_CST we_PPIS2 propose_VV0 ._. 
</s>
<s>
The_AT speedups_NN2 are_VBR shown_VVN to_TO be_VBI partly_RR due_II21 to_II22 the_AT novel_JJ search_NN1 strategy_NN1 and_CC partly_RR due_II21 to_II22 the_AT asynchrony_NN1 and_CC parallelism_NN1 allowed_VVN by_II the_AT search_NN1 strategy_NN1 ._. 
</s>
<s>
Fig._NN1 8_MC plots_NN2 the_AT resulting_JJ modularity_NN1 values_NN2 ._. 
</s>
<s>
On_II the_AT other_JJ hand_NN1 ,_, the_AT expressions_NN2 for_IF Cmin_NP1 and_CC C1_FO are_VBR satisfied_VVN even_RR for_IF the_AT simple_JJ case_NN1 where_RRQ opacities_NN2 appear_VV0 in_II the_AT upper_JJ parts_NN2 of_IO the_AT lung_NN1 ._. 
</s>
<s>
Given_VVN the_AT large_JJ numbers_NN2 of_IO triples_NN2 involved_VVD ,_, this_DD1 can_VM be_VBI difficult_JJ ._. 
</s>
<s>
For_REX21 instance_REX22 ,_, the_AT term_NN1 Natalia_NP1 sometimes_RT refers_VVZ to_II the_AT musician_NN1 ,_, but_CCB ordinarily_RR denotes_VVZ simply_RR a_AT1 person_NN1 '_NULL s_ZZ1 first_MD name_NN1 ,_, which_DDQ has_VHZ no_AT entry_NN1 in_II the_AT taxonomy_NN1 ._. 
</s>
<s>
Although_CS the_AT RF_NP1 algorithms_NN2 provide_VV0 a_AT1 performance_NN1 boost_NN1 ,_, one_MC1 of_IO the_AT major_JJ drawbacks_NN2 is_VBZ that_CST the_AT users_NN2 do_VD0 not_XX always_RR provide_VVI enough_DD feedback_NN1 information_NN1 to_TO make_VVI the_AT methods_NN2 perform_VVI effectively_RR ._. 
</s>
<s>
Finally_RR ,_, the_AT dash-dotted_JJ line_NN1 corresponds_VVZ to_II our_APPGE stress-free_JJ spectrin-level_JJ model_NN1 with_IW Nv=27,344_FO ._. 
</s>
<s>
To_TO counter_VVI the_AT hacker_NN1 ,_, the_AT government_NN1 planted_VVN a_AT1 decoy_NN1 ZIP_NN1 archive_NN1 on_II one_MC1 of_IO its_APPGE infected_JJ machines_NN2 ,_, which_DDQ the_AT malware_NN1 dutifully_RR exfiltrated_VVD to_II the_AT drop_NN1 site_NN1 ._. 
</s>
<s>
Almost_RR all_DB of_IO them_PPHO2 use_VV0 the_AT orthogonal_JJ variability_NN1 modeling_NN1 (_( OVM_NP1 )_) proposed_VVD by_II Pohl_NP1 et_RA21 al_RA22 ._. 
</s>
<s>
in_II &lsqb;_( 3_MC &rsqb;_) to_TO describe_VVI the_AT SPL_NP1 variability_NN1 without_IW extending_VVG its_APPGE architectural_JJ models_NN2 with_IW new_JJ notations_NN2 ._. 
</s>
<s>
For_IF individuals_NN2 who_PNQS used_VMK to_TO have_VHI an_AT1 exercise_NN1 regimen_NN1 ,_, being_VBG obese_JJ was_VBDZ the_AT highest_JJT predictor_NN1 of_IO low_JJ physical_JJ fitness_NN1 (_( odds_NN2 ratio_NN1 of_IO 2.53_MC )_) ,_, compared_VVN to_II age_NN1 or_CC frequency_NN1 of_IO exercise_NN1 &lsqb;_( 79_MC &rsqb;_) ._. 
</s>
<s>
It_PPH1 is_VBZ to_TO be_VBI noted_VVN that_CST in_II reality_NN1 it_PPH1 may_VM not_XX be_VBI feasible_JJ to_TO jointly_RR protect_VVI all_DB connections_NN2 in_II the_AT network_NN1 ._. 
</s>
<s>
For_IF each_DD1 test_NN1 we_PPIS2 had_VHD 50_MC questions_NN2 ._. 
</s>
<s>
Under_II stable_JJ conditions_NN2 based_VVN on_II our_APPGE experiments_NN2 ,_, the_AT number_NN1 of_IO control_NN1 messages_NN2 sent_VVN by_II each_DD1 MSN_NP1 in_II a_AT1 transformation_NN1 period_NN1 (_( say_VV0 about_II 30s_MC2 )_) is_VBZ proportional_JJ to_II its_APPGE degree_NN1 in_II the_AT overlay_NN1 structure_NN1 ._. 
</s>
<s>
In_II the_AT first_MD approach_NN1 ,_, the_AT decision_NN1 is_VBZ made_VVN for_IF the_AT class_NN1 with_IW highest_JJT distance_NN1 to_II other_JJ classes_NN2 ._. 
</s>
<s>
The_AT reconstruction_NN1 error_NN1 used_VVN in_II this_DD1 paper_NN1 assumes_VVZ that_CST data_NN have_VH0 a_AT1 Gaussian_JJ distribution_NN1 in_II the_AT feature_NN1 space_NN1 ._. 
</s>
<s>
From_II a_AT1 Non-Forensic_JJ Stakeholder_NN1 '_NULL s_ZZ1 perspective_NN1 ,_, Forensic_JJ Training_NN1 is_VBZ conducted_VVN around_II awareness_NN1 :_: "_" <staff>_FO are_VBR going_VVGK to_TO need_VVI to_TO know_VVI what_DDQ the_AT processes_NN2 are_VBR if_CS things_NN2 happen_VV0 "_" Expert1_FO ._. 
</s>
<s>
However_RR ,_, these_DD2 individuals_NN2 also_RR exhibited_VVD higher_JJR levels_NN2 of_IO collective_JJ attitudes_NN2 (_( using_VVG higher_JJR amounts_NN2 of_IO "_" we_PPIS2 "_" ,_, "_" us_PPIO2 "_" ,_, "_" our_APPGE "_" )_) towards_II project_NN1 completion_NN1 (_( noted_VVN for_IF the_AT measures_NN2 for_IF Abbrev_NP1 ._. 
</s>
<s>
The_AT methodology_NN1 also_RR computes_VVZ carbon_NN1 emissions_NN2 from_II the_AT energy_NN1 used_VVN during_II the_AT network_NN1 '_NULL s_ZZ1 operation_NN1 ._. 
</s>
<s>
Memory-based_NN1 ,_, or_CC instance-based_NN1 ,_, machine_VV0 learning_VVG techniques_NN2 classify_VV0 incoming_JJ emails_NN2 according_II21 to_II22 their_APPGE similarity_NN1 to_II stored_JJ examples_NN2 (_( i.e._REX training_NN1 emails_NN2 )_) ._. 
</s>
<s>
One_MC1 of_IO the_AT reasons_NN2 for_IF this_DD1 is_VBZ that_DD1 simple_JJ methods_NN2 exist_VV0 for_IF KDE_NN1 that_CST are_VBR a_AT1 function_NN1 of_IO only_RR a_AT1 single_JJ parameter_NN1 ,_, the_AT kernel_NN1 bandwidth_NN1 ,_, otherwise_RR termed_VVD the_AT smoothing_JJ parameter_NN1 (_( Scott_NP1 ,_, 1992_MC ,_, Wand_NN1 and_CC Jones_NP1 ,_, 1995_MC )_) ._. 
</s>
<s>
Two-thirds_MF of_IO the_AT primary_JJ studies_NN2 have_VH0 explicitly_RR considered_VVN threats_NN2 to_II validity_NN1 of_IO study_NN1 design_NN1 ._. 
</s>
<s>
Modeling_VVG such_DA a_AT1 system_NN1 with_IW a_AT1 standard_JJ network_NN1 flow_NN1 model_NN1 results_NN2 in_II nonlinear_JJ constraints_NN2 because_CS the_AT waste_NN1 materials_NN2 enter_VV0 the_AT system_NN1 as_II a_AT1 mixed_JJ entity_NN1 ,_, which_DDQ reflects_VVZ the_AT waste_NN1 composition_NN1 of_IO the_AT generated_JJ mixed_JJ waste_NN1 ._. 
</s>
<s>
In_II tandem_NN1 ,_, significant_JJ effort_NN1 must_VM be_VBI given_VVN to_II developing_VVG compelling_JJ interfaces_NN2 able_JK to_TO display_VVI structured_JJ data_NN from_II across_II the_AT Web_NN1 ._. 
</s>
<s>
Also_RR ,_, the_AT results_NN2 from_II the_AT experiment_NN1 helped_VVD us_PPIO2 revise_VVI the_AT NLtoSTD-BB_JJ translation_NN1 process_NN1 to_TO help_VVI improve_VVI its_APPGE effectiveness_NN1 when_CS applied_VVN on_II NL_NP1 &lsqb;_( 37_MC &rsqb;_) ._. 
</s>
<s>
Preventing_VVG unauthorized_JJ accesses_VVZ to_II memory_NN1 is_VBZ fundamental_JJ to_II both_DB2 effective_JJ debugging_NN1 ,_, error_NN1 prevention_NN1 ,_, and_CC computer_NN1 security_NN1 ._. 
</s>
<s>
A_AT1 wireless_NN1 network_NN1 has_VHZ a_AT1 somewhat_RR higher_RRR predicted_VVN perceived_JJ risk_NN1 than_CSN a_AT1 wired_JJ network_NN1 ._. 
</s>
<s>
One_MC1 important_JJ mechanism_NN1 for_IF protecting_VVG corporate_JJ information_NN1 ,_, in_II an_AT1 attempt_NN1 to_TO detect_VVI ,_, prevent_VV0 and_CC respond_VVI to_II security_NN1 breaches_NN2 is_VBZ through_II the_AT formulation_NN1 and_CC application_NN1 of_IO an_AT1 information_NN1 security_NN1 policy_NN1 (_( InSPy_NP1 )_) (_( Hone_NP1 and_CC Eloff_NP1 ,_, 2002andvon_FO Solms_VVZ and_CC von_NP1 Solms_NP1 ,_, 2004_MC )_) ._. 
</s>
<s>
In_II &lsqb;_( 14_MC &rsqb;_) ,_, Zhu_NP1 et_RA21 al_RA22 ._. 
</s>
<s>
proposed_VVN a_AT1 hand_NN1 gesture_NN1 recognition_NN1 system_NN1 for_IF HCI_NP1 ,_, where_CS a_AT1 hand_NN1 gesture_NN1 is_VBZ spotted_VVN based_VVN on_II the_AT temporal_JJ length_NN1 of_IO a_AT1 moving_JJ hand_NN1 in_II a_AT1 video_NN1 stream_NN1 ._. 
</s>
<s>
We_PPIS2 limit_VV0 possible_JJ approximations_NN2 to_II the_AT four_MC types_NN2 presented_VVN above_RL ._. 
</s>
<s>
The_AT spatially-enabled_JJ RDBMS_NP1 (_( i.e._REX ,_, PostgreSQL_VV0 with_IW PostGIS_NP1 extension_NN1 enabled_VVN )_) that_CST serves_VVZ as_II the_AT back-end_NN1 for_IF Ontop-spatial_JJ and_CC Strabon_NP1 ,_, on_II the_AT other_JJ hand_NN1 ,_, incorporates_VVZ more_RGR advanced_JJ and_CC mature_JJ techniques_NN2 for_IF efficient_JJ geospatial_JJ query_NN1 processing_NN1 which_DDQ are_VBR considered_VVN standard_JJ for_IF the_AT relational_JJ database_NN1 community_NN1 ,_, such_II21 as_II22 support_NN1 for_IF R-tree_JJ indices_NN2 and_CC spatially-enabled_JJ optimizer_NN1 ._. 
</s>
<s>
The_AT ?_? conjunctive_NN1 abstraction_NN1 ?_? techniques_NN2 in_II ?_? &lsqb;_( 30_MC &rsqb;_) ?_? were_VBDR an_AT1 initial_JJ version_NN1 of_IO that_DD1 approach_NN1 ._. 
</s>
<s>
Previous_JJ research_NN1 on_II DTNs_NP1 has_VHZ studied_VVN tradeoffs_NN2 between_II the_AT packet_NN1 delivery_NN1 delay_NN1 and_CC specific_JJ other_JJ metrics_NN2 &lsqb;_( 2_MC &rsqb;_) ,_, &lsqb;_( 3_MC &rsqb;_) ,_, &lsqb;_( 4_MC &rsqb;_) ,_, &lsqb;_( 5_MC &rsqb;_) ,_, &lsqb;_( 6_MC &rsqb;_) ,_, &lsqb;_( 7_MC &rsqb;_) and_CC &lsqb;_( 8_MC &rsqb;_) ._. 
</s>
<s>
No_AT modification_NN1 or_CC supplementary_JJ operation_NN1 is_VBZ necessary_JJ at_II intermediate_JJ recoding_NN1 since_CS the_AT proposed_JJ solution_NN1 is_VBZ homomorphic_JJ ._. 
</s>
<s>
For_REX21 example_REX22 ,_, Watson_NP1 carefully_RR followed_VVD its_APPGE statistical_JJ algorithm_NN1 to_TO proffer_VVI Toronto_NP1 as_II a_AT1 city_NN1 in_II the_AT United_NP1 States_NP1 ._. 
</s>
<s>
Once_RR a_AT1 hot_JJ spot_NN1 node_NN1 is_VBZ selected_VVN as_II a_AT1 target_NN1 for_IF test_NN1 generation_NN1 ,_, a_AT1 backward_JJ tracking_NN1 is_VBZ performed_VVN to_TO identify_VVI a_AT1 path_NN1 from_II the_AT selected_JJ node_NN1 to_II the_AT start_NN1 node_NN1 ._. 
</s>
<s>
On_II inputsYi_NN2 ,_, si_FW ,_, and_CC ,_, and_CC for_IF a_AT1 given_JJ random_JJ order-preserving_JJ functionf_NN1 ,_, it_PPH1 first_MD selects_VVZ a_AT1 randomsuch_NN1 thatsatisfies_NN2 the_AT information_NN1 thatincludes_VVZ aboutZiand_NN1 in_RR21 particular_RR22 the_AT order_NN1 relations_NN2 betweenZiand_VV0 elements_NN2 ofVi_NN2 ._. 
</s>
<s>
As_CSA one_PN1 can_VM see_VVI ,_, the_AT graph_NN1 has_VHZ been_VBN "_" cut_VVN open_JJ "_" at_II the_AT line_NN1 ≡_NULL 0_MC ._. 
</s>
<s>
Second_MD ,_, it_PPH1 can_VM be_VBI used_VVN as_II an_AT1 input_NN1 to_TO calculate_VVI the_AT effect_NN1 size_NN1 of_IO any_DD change_NN1 in_II accuracy_NN1 relative_II21 to_II22 chance_NN1 ._. 
</s>
<s>
If_CS x_ZZ1 is_VBZ not_XX the_AT root_NN1 of_IO G_ZZ1 ,_, then_RT to_TO set_VVI rW&lsqb;x&rsqb;_NNU or_CC prW&lsqb;x&rsqb;_NNU ,_, T_ZZ1 must_VM have_VHI a_AT1 prW_NNU or_CC priW_VV0 lock_NN1 on_II all_DB of_IO x_ZZ1 '_NULL s_ZZ1 parents_NN2 ._. 
</s>
<s>
This_DD1 allows_VVZ new_JJ complex_JJ workflows_NN2 and_CC scripting_VVG functionalities_NN2 to_TO be_VBI introduced_VVN quickly_RR to_TO extend_VVI AHE_NN1 ._. 
</s>
<s>
For_IF this_DD1 reason_NN1 ,_, we_PPIS2 limit_VV0 our_APPGE testing_NN1 set_VVN to_II the_AT most_RGT recent_JJ 300_MC issues_NN2 ._. 
</s>
<s>
The_AT paper_NN1 is_VBZ organized_VVN as_CSA follows_VVZ :_: Section_NN1 2_MC reviews_VVZ related_JJ work_NN1 identifying_VVG the_AT main_JJ gaps_NN2 in_II the_AT database_NN1 schema_NN1 evolution_NN1 and_CC the_AT associated_JJ application_NN1 code_NN1 testing_NN1 ._. 
</s>
<s>
The_AT biggest_JJT cluster_NN1 has_VHZ 50_MC 000_MC instances_NN2 ._. 
</s>
<s>
Significant_JJ value_NN1 can_VM also_RR be_VBI drawn_VVN from_II considering_VVG dumb_JJ possessions_NN2 that_CST we_PPIS2 might_VM not_XX readily_RR expect_VVI ,_, especially_RR those_DD2 that_CST are_VBR not_XX immediately_RR visible_JJ but_CCB are_VBR carried_VVN on_RP a_AT1 daily_JJ basis_NN1 ._. 
</s>
<s>
Madougou_NN1 and_CC coworkers_NN2 propose_VV0 in_RP &lsqb;_( 29_MC &rsqb;_) a_AT1 provenance-based_JJ approach_NN1 aimed_VVN at_II analyzing_VVG the_AT e-BioInfra_JJ e-Science_JJ platform_NN1 usage_NN1 and_CC identifying_VVG the_AT causes_NN2 of_IO application_NN1 failures_NN2 ._. 
</s>
<s>
Groundwater_NN1 is_VBZ modelled_VVN as_II a_AT1 store_NN1 which_DDQ is_VBZ recharged_VVN when_CS the_AT level_NN1 in_II the_AT pervious_JJ soil_NN1 store_NN1 exceeds_VVZ the_AT field_NN1 capacity_NN1 ._. 
</s>
<s>
Fig._NN1 11_MC compares_VVZ the_AT classification_NN1 errors_NN2 of_IO face_NN1 recognition_NN1 among_II four_MC different_JJ representation_NN1 methods_NN2 such_II21 as_II22 CLBP_NP1 ,_, ULBP_NP1 ,_, LBP_NP1 ,_, and_CC MCT_NP1 using_VVG the_AT TYPE1_FO probe_NN1 set_NN1 ,_, when_CS the_AT number_NN1 of_IO CLBP_NP1 codes_NN2 changes_NN2 ._. 
</s>
<s>
It_PPH1 is_VBZ to_TO be_VBI expected_VVN that_CST many_DA2 features_NN2 that_CST are_VBR shared_VVN will_VM be_VBI unique_JJ to_II that_DD1 pair_NN of_IO images_NN2 ._. 
</s>
<s>
The_AT algorithm_NN1 then_RT creates_VVZ a_AT1 new_JJ set_NN1 of_IO detection_NN1 modules_NN2 in_II a_AT1 random_JJ order_NN1 denoted_VVD by_RP in_II Line_NN1 3_MC ._. 
</s>
<s>
The_AT whole_JJ graph_NN1 has_VHZ 52_MC goals_NN2 ,_, 32_MC of_IO which_DDQ are_VBR leaf_NN1 goals_NN2 ._. 
</s>
<s>
As_II a_AT1 validation_NN1 step_NN1 ,_, four_MC test_NN1 cases_NN2 were_VBDR used_VVN to_TO ensure_VVI that_CST predictions_NN2 from_II the_AT model_NN1 were_VBDR realistic_JJ ._. 
</s>
<s>
For_IF all_DB i_ZZ1 ,_, the_AT truth_NN1 values_NN2 of_IO q_ZZ1 (_( y1_FO ,_, xi_NN1 )_) ,_, q_ZZ1 (_( xi_NN1 ,_, y1_FO )_) ,_, q_ZZ1 (_( y2_FO ,_, xi_NN1 )_) andq_NN1 (_( xi_NN1 ,_, y2_FO )_) are_VBR the_AT same_DA ._. 
</s>
<s>
Difficulty_NN1 increases_VVZ with_IW z_ZZ1 ,_, reflecting_VVG Lemma_NN1 22_MC ._. 
</s>
<s>
The_AT intuitive_JJ presentation_NN1 of_IO this_DD1 metadata_NN1 is_VBZ crucial_JJ to_TO truly_RR take_VVI advantage_NN1 of_IO its_APPGE value_NN1 ._. 
</s>
<s>
The_AT brain_NN1 processes_NN2 information_NN1 both_RR rationally_RR and_CC emotionally_RR ,_, although_CS emotions_NN2 about_II rational_JJ content_NN1 are_VBR usually_RR processed_VVN by_II the_AT brain_NN1 split_JJ seconds_NNT2 before_II rational_JJ or_CC logical_JJ interventions_NN2 by_II the_AT cerebral_JJ cortex_NN1 ._. 
</s>
<s>
In_II real_JJ implementation_NN1 ,_, a_AT1 source_NN1 would_VM have_VHI data_NN to_TO be_VBI sent_VVN out_RP ,_, say_VV0 ?_? m1_FO ,_, m2_FO ,_, ?_? 
</s>
<s>
The_AT buffering_JJ model_NN1 divides_VVZ the_AT packet_NN1 streams_NN2 into_II generations_NN2 ,_, each_DD1 of_IO size_NN1 ?_? h_ZZ1 ,_, such_CS21 that_CS22 the_AT packets_NN2 of_IO the_AT same_DA generation_NN1 are_VBR tagged_VVN with_IW a_AT1 common_JJ generation_NN1 number_NN1 ?_? 
</s>
<s>
NG_NP1 ?_? ._. 
</s>
<s>
There_EX may_VM be_VBI a_AT1 delay_NN1 ,_, but_CCB at_II some_DD point_NN1 the_AT entrants_NN2 '_NULL new_JJ business_NN1 model_NN1 and_CC innovative_JJ use_NN1 of_IO technology_NN1 will_VM force_VVI the_AT incumbent_NN1 to_TO radically_RR modify_VVI its_APPGE model_NN1 ,_, or_CC to_TO adopt_VVI the_AT entrants_NN2 '_NULL model_NN1 in_BCL21 order_BCL22 to_TO survive_VVI ._. 
</s>
<s>
With_IW SemTag_NP1 '_NULL s_ZZ1 current_JJ shallow_JJ level_NN1 of_IO understanding_NN1 ,_, RDFS_NP1 &lsqb;_( 8_MC &rsqb;_) provides_VVZ an_AT1 adequate_JJ language_NN1 for_IF representing_VVG the_AT annotations_NN2 it_PPH1 generates_VVZ ._. 
</s>
<s>
For_IF most_DAT of_IO the_AT past_JJ 25_MC years_NNT2 ,_, points_VVZ in_II this_DD1 space_NN1 have_VH0 been_VBN roughly_RR correlated_VVN as_RR21 well_RR22 ,_, suggesting_VVG there_EX is_VBZ no_AT free_JJ lunch_NN1 :_: to_TO get_VVI better_JJR reasoning_NN1 ,_, you_PPY need_VV0 to_TO gather_VVI better_JJR data_NN and_CC represent_VVI it_PPH1 in_II a_AT1 more_RGR complex_JJ way_NN1 ._. 
</s>
<s>
The_AT figure_NN1 represents_VVZ a_AT1 parameter_NN1 sweep_NN1 of_IO the_AT number_NN1 of_IO BGP_NP1 streams_NN2 and_CC the_AT duration_NN1 of_IO the_AT history_NN1 windowh_NN1 ._. 
</s>
<s>
Another_DD1 aspect_NN1 that_CST seems_VVZ rather_RG unintuitive_JJ is_VBZ that_CST ,_, in_II31 addition_II32 to_II33 initially_RR setting_VVG passwords_NN2 via_II this_DD1 route_NN1 ,_, users_NN2 must_VM also_RR go_VVI to_II the_AT '_NULL Save_VV0 As_CSA '_NULL dialog_NN1 box_NN1 in_BCL21 order_BCL22 to_TO change_VVI or_CC remove_VVI them_PPHO2 ._. 
</s>
<s>
Feedback_NN1 from_II seminars_NN2 and_CC conferences_NN2 enabled_VVD the_AT investigators_NN2 a_AT1 period_NN1 of_IO reflection_NN1 before_II drawing_VVG final_JJ conclusions_NN2 on_II recommendations_NN2 for_IF industry_NN1 ,_, contribution_NN1 to_II literature_NN1 ,_, and_CC proposals_NN2 for_IF further_JJR research_NN1 ._. 
</s>
<s>
The_AT sum_NN1 of_IO the_AT values_NN2 over_II each_DD1 edge_NN1 forms_VVZ an_AT1 approximation_NN1 to_II summing_VVG over_II Eq_NN1 ._. 
</s>
<s>
(_( 20_MC )_) ._. 
</s>
<s>
R-table_NN1 contains_VVZ the_AT location_NN1 of_IO the_AT hypothetical_JJ center_NN1 of_IO the_AT seal_NN1 model_NN1 ._. 
</s>
<s>
In_II other_JJ words_NN2 ,_, appropriately_RR grouped_VVN IVPs_NN2 would_VM take_VVI similar_JJ numbers_NN2 of_IO steps_NN2 ._. 
</s>
<s>
For_IF ?_? 
</s>
<s>
X=64_FO ,_, ?_? results_VVZ are_VBR analyzed_VVN for_IF the_AT four_MC strategies_NN2 ,_, and_CC the_AT worst_JJT case_NN1 (_( TopologyT2M1_FO ,_, ?_? 
</s>
<s>
Fig._NN1 ?_? 7(a)–_NN1 (_( c_ZZ1 )_) )_) ,_, intermediate_JJ case_NN1 (_( Topology_NN1 ?_? 
</s>
<s>
T2M4_FO ,_, ?_? 
</s>
<s>
Fig._NN1 ?_? 7(d)–_NN1 (_( f_ZZ1 )_) )_) and_CC the_AT best_JJT case_NN1 (_( Topology_NN1 ?_? 
</s>
<s>
T1M4_FO ,_, ?_? 
</s>
<s>
Fig._NN1 ?_? 7(g)–_NN1 (_( i_ZZ1 )_) )_) are_VBR shown_VVN ._. 
</s>
<s>
No_AT task_NN1 required_JJ access_NN1 to_II the_AT properties_NN2 tabs_NN2 ._. 
</s>
<s>
The_AT goal_NN1 is_VBZ only_RR to_TO remove_VVI a_AT1 triple_NN1 ._. 
</s>
<s>
The_AT middle_JJ point_NN1 or_CC pivot_NN1 of_IO the_AT interval_NN1 will_VM serve_VVI as_II the_AT comparative_JJ point_NN1 between_II both_DB2 measures_NN2 ._. 
</s>
<s>
The_AT node_NN1 sets_VVZ the_AT next-hop_JJ address_NN1 in_II the_AT routing_NN1 packet_NN1 '_NULL s_ZZ1 encapsulating_JJ header_NN1 (_( e.g._REX IP_NP1 header_NN1 destination_NN1 address_NN1 ?_? &lsqb;_( 16_MC &rsqb;_) )_) and_CC attempts_NN2 to_TO send_VVI the_AT packet_NN1 ._. 
</s>
<s>
This_DD1 theory_NN1 has_VHZ been_VBN used_VVN in_II physics_NN1 ,_, complexity_NN1 science_NN1 ,_, theoretical_JJ biology_NN1 ,_, microstructure_VV0 modeling_NN1 ,_, and_CC spatial_JJ modeling_NN1 ._. 
</s>
<s>
The_AT Agricultural_JJ Production_NN1 module_NN1 allows_VVZ users_NN2 to_TO examine_VVI projected_JJ agricultural_JJ production_NN1 under_II recent_JJ historical_JJ (_( termed_VVN current_JJ )_) climate_NN1 and_CC three_MC potential_JJ warming_NN1 and_CC drying_VVG scenarios_NN2 (_( Table_NN1 ?_? 1_MC1 )_) ,_, as_II31 well_II32 as_II33 four_MC commodity_NN1 price_NN1 scenarios_NN2 and_CC four_MC production_NN1 cost_NN1 scenarios_NN2 (_( Table_NN1 ?_? 2_MC )_) ._. 
</s>
<s>
Typically_RR ,_, intrusion_NN1 detection_NN1 refers_VVZ to_II a_AT1 variety_NN1 of_IO techniques_NN2 for_IF detecting_VVG attacks_NN2 in_II the_AT form_NN1 of_IO malicious_JJ and_CC unauthorized_JJ activities_NN2 ._. 
</s>
<s>
The_AT obvious_JJ choice_NN1 for_IF terms_NN2 in_II text_NN1 documents_NN2 are_VBR the_AT words_NN2 that_CST are_VBR directly_RR related_VVN to_II the_AT semantic_JJ contents_NN2 ._. 
</s>
<s>
First_MD in_II Section_NN1 3.3.1_MC we_PPIS2 present_VV0 the_AT construction_NN1 that_CST given_VVN a_AT1 POMDP_NN1 G_ZZ1 constructs_NN2 an_AT1 exponential_NN1 size_NN1 belief-observation_NN1 POMDP_NN1 such_CS21 that_CS22 there_EX exists_VVZ a_AT1 finite-memory_JJ almost-sure_JJ winning_JJ strategy_NN1 for_IF the_AT objective_NN1 LimAvg=1_FO in_II G_ZZ1 iff_CS there_EX exists_VVZ a_AT1 randomized_JJ memoryless_JJ almost-sure_JJ winning_JJ strategy_NN1 in_RP for_IF the_AT LimAvg=1objective_FO ._. 
</s>
<s>
Considering_CS the_AT distinct_JJ set_NN1 of_IO IDS_NN2 functions_NN2 on_II each_DD1 node_NN1 along_II a_AT1 routing_NN1 path_NN1 ,_, the_AT entire_JJ traffic_NN1 on_II that_DD1 path_NN1 is_VBZ investigated_VVN by_II more_DAR IDS_NN2 functions_NN2 while_CS none_PN of_IO the_AT nodes_NN2 is_VBZ overloaded_VVN ._. 
</s>
<s>
Life_NN1 cycle_NN1 impact_NN1 factors_NN2 can_VM then_RT be_VBI used_VVN with_IW the_AT life_NN1 cycle_NN1 inventory_NN1 results_NN2 to_TO calculate_VVI environmental_JJ impacts_NN2 from_II the_AT emissions_NN2 (_( e.g._REX ,_, global_JJ warming_NN1 potential_NN1 ,_, acidification_NN1 potential_NN1 ,_, or_CC human_JJ toxicity_NN1 )_) ._. 
</s>
<s>
We_PPIS2 write_VV0 it_PPH1 as_II a_AT1 matrix-valued_JJ function_NN1 with_IW three_MC columns_NN2 ,_, However_RR ,_, even_RR in_II its_APPGE initial_JJ state_NN1 ,_, the_AT coupled_JJ code_NN1 is_VBZ capable_JJ of_IO handling_VVG large_JJ systems_NN2 containing_JJ tens_MC2 of_IO millions_NNO2 of_IO dislocation_NN1 segments_NN2 and_CC executing_VVG on_II many_DA2 processors_NN2 with_IW good_JJ performance_NN1 ._. 
</s>
<s>
Core_NN1 developers_NN2 '_NULL communication_NN1 networks_NN2 are_VBR likely_JJ to_TO be_VBI less_RGR dense_JJ and_CC their_APPGE closeness_NN1 measures_VVZ higher_JJR (_( noting_VVG that_CST higher_JJR measures_NN2 for_IF network_NN1 closeness_NN1 indicate_VV0 that_CST members_NN2 are_VBR less_RGR reachable_JJ )_) during_II the_AT latter_DA project_NN1 phase_NN1 ,_, as_CSA per_II their_APPGE reduction_NN1 in_II communication_NN1 as_CSA noted_VVN in_II Table_NN1 4_MC ._. 
</s>
<s>
Williams_NP1 et_RA21 al_RA22 ._. 
</s>
<s>
&lsqb;_( 8_MC &rsqb;_) presented_VVD a_AT1 hierarchical_JJ autotuning_JJ model_NN1 for_IF parallel_JJ lattice-Boltzmann_NP1 ,_, and_CC report_VV0 a_AT1 performance_NN1 increase_NN1 of_IO more_DAR than_CSN a_AT1 factor_NN1 3_MC in_II their_APPGE simulations_NN2 ._. 
</s>
<s>
Slight_JJ modifications_NN2 to_II the_AT tool_NN1 were_VBDR required_VVN to_TO ensure_VVI that_CST the_AT file_NN1 names_NN2 produced_VVN in_II the_AT changesets_NN2 included_VVD the_AT path_NN1 information_NN1 to_TO match_VVI the_AT file_NN1 names_NN2 produced_VVN by_II our_APPGE rlog_NN1 processing_NN1 script_NN1 ._. 
</s>
<s>
As_II21 for_II22 individual_JJ key_NN1 ,_, it_PPH1 is_VBZ only_RR shared_VVN by_II the_AT GC_NP1 and_CC individual_JJ member_NN1 ._. 
</s>
<s>
However_RR ,_, the_AT Provenance_NN1 Working_JJ Group_NN1 felt_VVD that_CST both_DB2 use_VV0 cases_NN2 were_VBDR important_JJ :_: indeed_RR ,_, many_DA2 existing_JJ provenance_NN1 systems_NN2 already_RR support_VV0 precise_JJ capture_NN1 of_IO provenance_NN1 ,_, whereas_CS enabling_VVG Web_NN1 pages_NN2 to_TO be_VBI marked_VVN up_RP using_VVG prov_NN1 was_VBDZ a_AT1 key_JJ part_NN1 of_IO why_RRQ the_AT working_JJ group_NN1 was_VBDZ chartered_VVN ._. 
</s>
<s>
This_DD1 indicates_VVZ that_CST algorithmic_JJ teaching_NN1 guidance_NN1 is_VBZ preferable_JJ over_II heuristic_JJ when_CS it_PPH1 is_VBZ available_JJ ._. 
</s>
<s>
As_II a_AT1 final_JJ aspect_NN1 of_IO their_APPGE own_DA IT_NN1 usage_NN1 ,_, the_AT respondents_NN2 were_VBDR asked_VVN to_TO indicate_VVI whether_CSW they_PPHS2 used_VVD the_AT operating_NN1 system_NN1 and_CC applications_NN2 referenced_VVD later_RRR in_II the_AT survey_NN1 ._. 
</s>
<s>
In_RR21 addition_RR22 ,_, drift_NN1 and_CC decreased_VVD signal-to-noise_JJ ratio_NN1 can_VM occur_VVI over_II time_NNT1 as_II a_AT1 result_NN1 of_IO sensor_NN1 degradation_NN1 or_CC encapsulation_NN1 caused_VVN by_II an_AT1 inflammation_NN1 response_NN1 in_II the_AT microenvironment_NN1 around_II the_AT sensor_NN1 &lsqb;_( 8_MC &rsqb;_) ,_, &lsqb;_( 33_MC &rsqb;_) ,_, &lsqb;_( 34_MC &rsqb;_) and_CC &lsqb;_( 35_MC &rsqb;_) ._. 
</s>
<s>
The_AT goal_NN1 is_VBZ to_TO give_VVI the_AT chance_NN1 to_II the_AT new_JJ potential_JJ frequent_JJ IP_NN1 sets_VVZ to_TO replace_VVI the_AT existing_JJ ones_NN2 ._. 
</s>
<s>
His_APPGE concern_NN1 is_VBZ to_TO develop_VVI a_AT1 theoretical_JJ foundation_NN1 for_IF agile_JJ methods_NN2 ,_, that_REX21 is_REX22 ,_, an_AT1 agile_JJ method_NN1 rooted_VVN in_II practices_NN2 that_CST are_VBR known_VVN to_TO increase_VVI agility_NN1 and_CC hence_RR perform_VVI well_RR in_II uncertain_JJ situations_NN2 ._. 
</s>
<s>
The_AT number_NN1 of_IO search_NN1 networks_NN2 that_CST can_VM be_VBI represented_VVN by_II SIL_NP1 is_VBZ quite_RG large_JJ ._. 
</s>
<s>
This_DD1 section_NN1 presents_VVZ construction_NN1 of_IO a_AT1 probabilistic_JJ finite_JJ state_NN1 automaton_NN1 (_( PFSA_NP1 )_) for_IF feature_NN1 extraction_NN1 based_VVN on_II the_AT symbol_NN1 image_NN1 generated_VVN from_II a_AT1 wavelet_NN1 surface_NN1 profile_NN1 ._. 
</s>
<s>
The_AT intensity_NN1 of_IO the_AT point_NN1 represents_VVZ the_AT number_NN1 of_IO experiments_NN2 ,_, in_II a_AT1 series_NN of_IO random_JJ experiments_NN2 ,_, where_CS we_PPIS2 obtain_VV0 this_DD1 specific_JJ correlation_NN1 value_NN1 for_IF the_AT given_JJ ._. 
</s>
<s>
Clamping_NP1 is_VBZ a_AT1 usual_JJ and_CC useful_JJ tool_NN1 when_CS one_PN1 wants_VVZ to_TO reduce_VVI the_AT number_NN1 of_IO free_JJ parameters_NN2 of_IO a_AT1 model_NN1 in_II order_NN1 to_TO either_RR obtain_VVI a_AT1 better_JJR estimate_NN1 or_CC reduce_VV0 the_AT computational_JJ load_NN1 ._. 
</s>
<s>
When_CS using_VVG this_DD1 technique_NN1 ,_, concepts_NN2 such_II21 as_II22 B_ZZ1 are_VBR called_VVN pseudo_JJ nominals_NN2 ._. 
</s>
<s>
However_RR ,_, a_AT1 preliminary_JJ experimentation_NN1 suggested_VVD that_CST this_DD1 is_VBZ not_XX the_AT right_JJ way_NN1 either_RR ._. 
</s>
<s>
Although_CS it_PPH1 has_VHZ been_VBN shown_VVN that_CST such_DA hijacks_NN2 could_VM be_VBI used_VVN for_IF sending_VVG spam_NN1 from_II unused_JJ address_NN1 space&lsqb;19&rsqb;_FO ,_, it_PPH1 could_VM not_XX be_VBI used_VVN to_TO divert_VVI traffic_NN1 away_II21 from_II22 proper_JJ destinations_NN2 because_CS routers_NN2 always_RR forward_JJ packets_NN2 to_II the_AT smallest_JJT matching_JJ prefix_NN1 ,_, and_CC in_II this_DD1 paper_NN1 we_PPIS2 do_VD0 not_XX consider_VVI such_DA attacks_NN2 ._. 
</s>
<s>
Since_CS Twitter_NN1 data_NN are_VBR no_RR21 longer_RR22 available_JJ to_II researchers_NN2 ,_, this_DD1 remains_VVZ the_AT largest_JJT available_JJ snapshot_NN1 of_IO Twitter_NN1 ,_, with_IW 41_MC million_NNO user_NN1 profiles_NN2 and_CC 1.47_MC billion_NNO social_JJ relations_NN2 ._. 
</s>
<s>
Based_VVN on_II the_AT literature_NN1 &lsqb;_( 61_MC &rsqb;_) ,_, we_PPIS2 expected_VVD lower_JJR privacy_NN1 concerns_VVZ from_II Group_NN1 India_NP1 than_CSN Group_NN1 US_NP1 ._. 
</s>
<s>
As_CSA in_II the_AT case_NN1 with_IW the_AT Eclipse-Classic_JJ product_NN1 ,_, we_PPIS2 observe_VV0 that_CST across_II the_AT product_NN1 line_NN1 ,_, results_NN2 show_VV0 an_AT1 improving_JJ trend_NN1 for_IF all_DB products_NN2 in_II the_AT UseAllPredictAll_NP1 and_CC UseAllPredictPost_VV0 datasets_NN2 ._. 
</s>
<s>
Input_NN1 variable_NN1 selection_NN1 (_( IVS_MC2 )_) ,_, as_CSA one_MC1 of_IO the_AT most_RGT important_JJ steps_NN2 in_II the_AT development_NN1 of_IO ANN_NP1 and_CC other_JJ data_NN driven_JJ environmental_JJ and_CC water_NN1 resources_NN2 models_NN2 ,_, determines_VVZ the_AT quality_NN1 and_CC quantity_NN1 of_IO information_NN1 used_VVN in_II the_AT modelling_NN1 process_NN1 ._. 
</s>
<s>
Efferent_JJ coupling_NN1 is_VBZ the_AT number_NN1 of_IO classes_NN2 which_DDQ make_VV0 reference_NN1 to_II a_AT1 class_NN1 and_CC can_VM be_VBI thought_VVN of_IO as_CSA the_AT number_NN1 of_IO uses_NN2 "_" outside_RL "_" of_IO the_AT class_NN1 ._. 
</s>
<s>
Ultimately_RR ,_, the_AT PE_NN1 pro-actively_RR monitors_VVZ channels_NN2 the_AT SSR_NP1 is_VBZ attempting_VVG to_TO use_VVI and_CC pro-actively_RR enforces_VVZ that_CST transmissions_NN2 originating_VVG from_II the_AT SSR_NP1 fully_RR satisfy_VV0 policy_NN1 requirements_NN2 ._. 
</s>
<s>
In_II this_DD1 problem_NN1 ,_, we_PPIS2 have_VH0 called_VVN the_AT presumed_JJ common_JJ invariant_JJ feature_NN1 "_" mutual_JJ face_NN1 "_" ._. 
</s>
<s>
One_MC1 of_IO the_AT most_RGT important_JJ examples_NN2 of_IO a_AT1 special_JJ purpose_NN1 operating_NN1 environment_NN1 is_VBZ the_AT sort_NN1 of_IO programmable_JJ logic_NN1 controller_NN1 used_VMK to_TO limit_VVI the_AT operation_NN1 of_IO devices_NN2 in_II some_DD infrastructure_NN1 systems_NN2 ,_, mechanisms_NN2 that_CST control_VV0 movements_NN2 of_IO physical_JJ devices_NN2 ,_, doses_NN2 of_IO radiation_NN1 ,_, and_RR31 so_RR32 forth_RR33 ._. 
</s>
<s>
The_AT drawback_NN1 of_IO this_DD1 model_NN1 is_VBZ that_CST the_AT model_NN1 coefficient_NN1 is_VBZ kept_VVN constant_JJ ,_, while_CS in_II reality_NN1 it_PPH1 should_VM vary_VVI within_II the_AT flow_NN1 field_NN1 depending_II21 on_II22 the_AT local_JJ state_NN1 of_IO turbulence_NN1 ._. 
</s>
<s>
We_PPIS2 distinguish_VV0 unsupervised_JJ ,_, weakly_RR supervised_VVN and_CC supervised_VVN approaches_NN2 ._. 
</s>
<s>
We_PPIS2 have_VH0 generated_VVN topologies_NN2 by_II reducing_VVG the_AT Internet_NN1 instance_NN1 I010507_FO using_VVG DHYB-0.8_MC and_CC DRV_NP1 ._. 
</s>
<s>
This_DD1 finding_NN1 could_VM have_VHI implications_NN2 for_IF aiding_VVG the_AT user_NN1 search_NN1 process_NN1 in_II a_AT1 variety_NN1 of_IO situations_NN2 that_CST involves_VVZ multi-faceted_JJ and_CC broad_JJ search_NN1 tasks_NN2 ._. 
</s>
<s>
Metrics_NN2 such_II21 as_II22 the_AT Shodan_NP1 survey_NN1 enable_VV0 us_PPIO2 to_TO test_VVI this_DD1 ._. 
</s>
<s>
So_RR ,_, although_CS finding_VVG the_AT structures_NN2 of_IO interest_NN1 for_IF a_AT1 specific_JJ implementation_NN1 and_CC version_NN1 of_IO Linux_NN1 may_VM indeed_RR be_VBI easier_JJR ,_, generalizing_VVG the_AT location_NN1 of_IO these_DD2 structures_NN2 across_II all_DB versions_NN2 of_IO Linux_NN1 to_TO produce_VVI a_AT1 single_JJ tool_NN1 that_CST can_VM find_VVI these_DD2 structures_NN2 in_II an_AT1 arbitrary_JJ Linux_NN1 memory_NN1 dump_NN1 is_VBZ not_XX feasible_JJ ._. 
</s>
<s>
We_PPIS2 developed_VVD a_AT1 computer_NN1 program_NN1 ,_, to_TO implement_VVI the_AT dynamic_JJ programming_NN1 computingh_NN1 (_( β_NULL )_) over_II a_AT1 range_NN1 of_IO values_NN2 of_IO β_NULL ,_, given_VVN a_AT1 Bernoulli_JJ arrival_NN1 distribution_NN1 with_IW a_AT1 packet_NN1 arrival_NN1 probability_NN1 ofpin_NN1 every_AT1 slot_NN1 ,_, a_AT1 fixed_JJ service_NN1 rate_NN1 of_IO one_MC1 packet_NN1 per_II slot_NN1 and_CC fixed_JJ power_NN1 costs_VVZ ofPa_NN1 ,_, Ps_NP1 ,_, Pas_FW ,_, Psa_NP1 ._. 
</s>
<s>
Instead_RR ,_, the_AT natural_JJ spacetime_NNT1 jump_NN1 condition_NN1 for_IF u_ZZ1 comes_VVZ from_II (_( 12a_FO )_) and(12b)_FO ,_, which_DDQ is_VBZ one_MC1 of_IO the_AT characteristic_JJ equations_NN2 of_IO the_AT elastodynamic_JJ system_NN1 ._. 
</s>
<s>
This_DD1 has_VHZ greater_JJR impact_NN1 on_II the_AT execution_NN1 time_NNT1 of_IO geospatial_JJ queries_NN2 ,_, as_CSA the_AT evaluation_NN1 of_IO spatial_JJ joins_NN2 is_VBZ more_RGR expensive_JJ due_II21 to_II22 the_AT cost_NN1 of_IO the_AT evaluation_NN1 of_IO the_AT spatial_JJ conditions_NN2 ._. 
</s>
<s>
The_AT percentiles_NN2 are_VBR meant_VVN to_TO give_VVI a_AT1 quantitative_JJ basis_NN1 for_IF comparison_NN1 between_II the_AT individual_JJ rating_NN1 levels_NN2 ._. 
</s>
<s>
This_DD1 can_VM be_VBI done_VDN on_II an_AT1 individual_JJ basis_NN1 ,_, or_CC with_IW all_DB participants_NN2 in_II a_AT1 group_NN1 meeting_NN1 ._. 
</s>
<s>
Procedures_NN2 for_IF quality_NN1 evaluation_NN1 of_IO SE_NP1 papers_NN2 when_RRQ the_AT primary_JJ studies_NN2 have_VH0 used_VVN a_AT1 variety_NN1 of_IO different_JJ empirical_JJ methods_NN2 ._. 
</s>
<s>
Correctly_RR designing_VVG a_AT1 system_NN1 that_CST relies_VVZ on_II a_AT1 set_NN1 of_IO complex_JJ security_NN1 policies_NN2 calls_VVZ for_IF a_AT1 new_JJ set_NN1 of_IO techniques_NN2 to_TO make_VVI it_PPH1 tractable_JJ for_IF a_AT1 human_NN1 to_TO correctly_RR formulate_VVI policy_NN1 specifications_NN2 ._. 
</s>
<s>
If_CS more_DAR than_CSN one_MC1 answers_VVZ are_VBR valid_JJ ,_, the_AT candidate_NN1 taking_VVG the_AT test_NN1 may_VM be_VBI asked_VVN to_TO pick_VVI the_AT best_RRT among_II them_PPHO2 ._. 
</s>
<s>
Since_CS the_AT processes_NN2 of_IO scholarly_JJ communication_NN1 are_VBR central_JJ to_II the_AT practice_NN1 of_IO science_NN1 ,_, it_PPH1 is_VBZ essential_JJ that_CST publishers_NN2 now_RT adopt_VV0 such_DA standards_NN2 to_TO permit_VVI inference_NN1 over_II the_AT entire_JJ corpus_NN1 of_IO scholarly_JJ communication_NN1 represented_VVN in_II journals_NN2 ,_, books_NN2 and_CC conference_NN1 proceedings_NN2 ._. 
</s>
<s>
These_DD2 are_VBR then_RT used_VVN to_TO observe_VVI that_DD1 formal_JJ methods_NN2 have_VH0 "_" not_XX seen_VVN widespread_JJ adoption?except_NN1 arguably_RR in_II the_AT development_NN1 of_IO critical_JJ systems_NN2 in_II certain_JJ domains_NN2 "_" ._. 
</s>
<s>
The_AT PseudoTag_NN1 recommender_NN1 system_NN1 is_VBZ also_RR able_JK to_TO provide_VVI a_AT1 more_RGR stable_JJ level_NN1 of_IO recommendation_NN1 quality_NN1 ._. 
</s>
<s>
This_DD1 step_NN1 is_VBZ necessary_JJ to_TO allow_VVI the_AT MH_NP1 to_TO communicate_VVI with_IW the_AT different_JJ entities_NN2 of_IO the_AT Space_NN1 ,_, maintain_VV0 presence_NN1 ,_, connect_VV0 to_II other_JJ users_NN2 within_II the_AT space_NN1 and_CC much_RR more_RRR ._. 
</s>
<s>
HTTP_NP1 traffic_NN1 with_IW less_DAR than_CSN 80%_NNU RFC_NP1 compliance_NN1 or_CC statistical_JJ deviations_NN2 are_VBR flagged_VVN as_CSA anomalies_NN2 ._. 
</s>
<s>
Similarly_RR ,_, the_AT SW_ND1 algorithm_NN1 conserved_VVD mass_NN1 up_II21 to_II22 the_AT imposed_JJ conservation_JJ tolerance_NN1 ,_, mc=10-6mc=10-6_FO ._. 
</s>
<s>
We_PPIS2 report_VV0 the_AT details_NN2 related_VVN to_II a_AT1 replicated_JJ study_NN1 separately_RR only_RR when_CS differences_NN2 exist_VV0 between_II the_AT original_JJ study_NN1 and_CC the_AT replicated_JJ study_NN1 in_II31 terms_II32 of_II33 a_AT1 particular_JJ study_NN1 design_NN1 aspect_NN1 ._. 
</s>
<s>
These_DD2 findings_NN2 ,_, along_II21 with_II22 other_JJ observations_NN2 ,_, have_VH0 led_VVN researchers_NN2 to_TO propose_VVI that_CST queries_NN2 can_VM be_VBI regarded_VVN as_II a_AT1 language_NN1 of_IO their_APPGE own_DA ,_, which_DDQ is_VBZ evolving_VVG at_II a_AT1 fast_JJ pace_NN1 &lsqb;_( 31_MC &rsqb;_) ,_, &lsqb;_( 25_MC &rsqb;_) ,_, &lsqb;_( 26_MC &rsqb;_) ,_, &lsqb;_( 32_MC &rsqb;_) ._. 
</s>
<s>
Eq_NN1 ._. 
</s>
<s>
(_( 13_MC )_) is_VBZ used_VVN to_TO determine_VVI the_AT available_JJ capacity_NN1 of_IO new_JJ treatment_NN1 processes_NN2 in_II the_AT initial_JJ stage_NN1 ._. 
</s>
<s>
The_AT Sinc-Collocation_NP1 Ekman_NP1 Spiral_JJ projection_NN1 of_IO Example_NN1 5_MC for_IF different_JJ values_NN2 of_IO N_ZZ1 against_II the_AT exact_JJ solution_NN1 while_CS =0_FO ,_, =45_FO ,_, =3.14_FO ,_, D0=60m_FO ,_, and_CC DE=19m_FO ._. 
</s>
<s>
In_II this_DD1 method_NN1 the_AT nodes_NN2 in_II the_AT protection_NN1 tree_NN1 are_VBR sorted_VVN in_II descending_JJ order_NN1 based_VVN on_II the_AT nodal_JJ degree_NN1 of_IO the_AT nodes_NN2 ._. 
</s>
<s>
To_TO clarify_VVI ,_, EliMet_NP1 did_VDD not_XX know_VVI about_II the_AT assigned_JJ security_NN1 measures_NN2 ,_, and_CC instead_RR ,_, estimated_VVD them_PPHO2 using_VVG the_AT optimal_JJ actions_NN2 reported_VVN by_II the_AT operator_NN1 simulator_NN1 ._. 
</s>
<s>
While_CS plain-text_JJ email_NN1 sent_VVN over_II the_AT Internet_NN1 can_VM hardly_RR be_VBI considered_VVN secure_JJ ,_, any_DD organisation_NN1 that_CST employs_VVZ an_AT1 off-site_JJ solution_NN1 is_VBZ likely_JJ to_TO have_VHI a_AT1 vested_JJ interest_NN1 in_II ensuring_VVG a_AT1 large_JJ repository_NN1 of_IO their_APPGE email_NN1 is_VBZ not_XX retained_VVN over_II time_NNT1 and_CC that_DD1 only_JJ authorised_JJ personnel_NN2 have_VH0 access_NN1 to_II what_DDQ could_VM be_VBI effectively_RR used_VVN as_II an_AT1 electronic_JJ wiretap_NN1 on_II their_APPGE email_NN1 communications_NN2 ._. 
</s>
<s>
Our_APPGE only_JJ deviation_NN1 from_II the_AT plan_NN1 occurred_VVD here_RL because_CS judgment_NN1 took_VVD 15_MC onths_NN2 after_II completion_NN1 of_IO the_AT court_NN1 proceedings_NN2 ._. 
</s>
<s>
By_II permitting_VVG this_DD1 limited_JJ misuse_NN1 of_IO the_AT device_NN1 ,_, it_PPH1 is_VBZ possible_JJ to_TO achieve_VVI a_AT1 much_RR higher_JJR level_NN1 of_IO user_NN1 convenience_NN1 at_II minimal_JJ expense_NN1 to_II the_AT security_NN1 ._. 
</s>
<s>
After_CS the_AT execution_NN1 of_IO the_AT algorithm_NN1 each_DD1 agent_NN1 knows_VVZ all_DB the_AT associations_NN2 of_IO its_APPGE planes_NN2 with_IW the_AT planes_NN2 of_IO the_AT rest_NN1 of_IO the_AT agents_NN2 ,_, even_CS21 if_CS22 they_PPHS2 are_VBR not_XX direct_JJ neighbors_NN2 ._. 
</s>
<s>
This_DD1 explains_VVZ why_RRQ its_APPGE runtime_NNT1 does_VDZ not_XX increase_VVI linearly_RR with_IW data_NN size_NN1 ._. 
</s>
<s>
The_AT Holm_NP1 procedure_NN1 adjusts_VVZ the_AT value_NN1 of_IO in_II a_AT1 step-down_JJ manner_NN1 ._. 
</s>
<s>
Despite_II these_DD2 possible_JJ shortcomings_NN2 ,_, there_EX is_VBZ a_AT1 high_JJ probability_NN1 that_CST the_AT proposed_JJ algorithm_NN1 will_VM improve_VVI upon_II the_AT triangulation_NN1 produced_VVN by_II the_AT unmodified_JJ CDT_NP1 technique_NN1 ._. 
</s>
<s>
Moreover_RR ,_, a_AT1 statistical_JJ approach_NN1 to_TO alert_VVI correlation_NN1 should_VM deal_VVI with_IW "_" concept_NN1 drift_NN1 "_" ._. 
</s>
<s>
The_AT results_NN2 of_IO level_JJ set_NN1 implementation_NN1 &lsqb;_( 3_MC &rsqb;_) in_II31 case_II32 of_II33 Fig.2_FO ,_, Fig.3_FO ,_, Fig.4_FO ,_, Fig.5_FO are_VBR shown_VVN in_II Figs._NN2 6(k)_FO ,_, (_( l_ZZ1 )_) ,_, (_( m_ZZ1 )_) ,_, (_( n_ZZ1 )_) ,_, (_( o_ZZ1 )_) ,_, (_( p_ZZ1 )_) and_CC (_( q_ZZ1 )_) ,_, respectively_RR ._. 
</s>
<s>
Link_VV0 capacities_NN2 are_VBR set_VVN up_RP depending_II21 on_II22 the_AT level_NN1 they_PPHS2 belong_VV0 to_II ._. 
</s>
<s>
There_EX is_VBZ no_AT indication_NN1 in_II that_DD1 document_NN1 that_CST computers_NN2 were_VBDR to_TO work_VVI together_RL with_IW people_NN ;_; the_AT intent_NN1 was_VBDZ to_TO develop_VVI an_AT1 autonomous_JJ intelligence_NN1 ._. 
</s>
<s>
For_IF small_JJ samples_NN2 ,_, the_AT Nelson–Aalen_JJ estimator_NN1 is_VBZ better_JJR than_CSN the_AT Kaplan–Meier_NN1 estimator_NN1 (_( Collett_NP1 ,_, 2003_MC ,_, p._NN1 22_MC )_) ._. 
</s>
<s>
A_AT1 uniform_JJ mesh_NN1 of_IO trilinear_JJ hexahedra_NN1 with_IW h=1/12h=1/12_FU was_VBDZ adopted_VVN for_IF all_DB examples_NN2 ._. 
</s>
<s>
The_AT world_NN1 is_VBZ a_AT1 global_JJ village_NN1 :_: the_AT processes_NN2 used_VMK to_TO produce_VVI the_AT goods_NN2 in_II China_NP1 may_VM affect_VVI the_AT future_JJ health_NN1 of_IO the_AT children_NN2 in_II America_NP1 (_( or_CC other_JJ nations_NN2 where_RRQ the_AT jewelry_NN1 was_VBDZ marketed_VVN and_CC sold_VVN )_) and_CC vice_RR21 versa_RR22 ._. 
</s>
<s>
In_II a_AT1 scenario_NN1 where_RRQ access_NN1 control_NN1 and_CC charging_NN1 is_VBZ primarily_RR based_VVN on_II 3GPP_FO systems_NN2 ,_, subscriber_NN1 authentication_NN1 is_VBZ handled_VVN by_II the_AT 3GPP_FO system_NN1 to_TO enhance_VVI WLAN-AN_AT1 security_NN1 ._. 
</s>
<s>
The_AT idea_NN1 of_IO using_VVG anomalies_NN2 as_CSA proxies_NN2 for_IF attacks_NN2 has_VHZ been_VBN extensively_RR studied_VVN in_II various_JJ security_NN1 domains_NN2 and_CC ,_, albeit_CS generally_RR useful_JJ ,_, is_VBZ not_XX free_JJ from_II drawbacks_NN2 and_CC controversies_NN2 (_( Sommer_NP1 and_CC Paxson_NP1 ,_, 2010_MC )_) ._. 
</s>
<s>
In_II AU331_FO the_AT APC_FO is_VBZ very_RG close_JJ to_II one_MC1 ,_, which_DDQ represents_VVZ a_AT1 matrix_NN1 without_IW movement_NN1 costs_NN2 ,_, and_CC reflects_VVZ well_RR the_AT more_RGR natural_JJ context_NN1 in_II this_DD1 sample_NN1 when_CS compared_VVN to_II AU113_FO ._. 
</s>
<s>
This_DD1 is_VBZ an_AT1 Open_JJ Geospatial_JJ Consortium_NN1 (_( OGC_NP1 )_) specification_NN1 for_IF how_RGQ geospatial_JJ inputs_NN2 and_CC outputs_NN2 should_VM be_VBI handled_VVN between_II different_JJ computer_NN1 software_NN1 units_NN2 ._. 
</s>
<s>
The_AT set_NN1 of_IO edgesEis_NN1 defined_VVN as_CSA follows_VVZ :_: for_IF every_AT1 organizationCwhose_NN1 delegation_NN1 policy_NN1 fory/kis_NN1 the_AT empty_JJ set_NN1 ,_, a_AT1 directed_JJ edge_NN1 is_VBZ placed_VVN betweenCand_NN1 ._. 
</s>
<s>
For_IF every_AT1 other_JJ organizationDand_NN1 every_AT1 pair_NN (_( y/k_ZZ1 ,_, Z_ZZ1 )_) inD_VV0 '_NULL s_ZZ1 delegation_NN1 policy_NN1 fory/k_FU ,_, a_AT1 directed_JJ edge_NN1 is_VBZ placed_VVN fromDtoZwhereZ_NN1 is_VBZ inOASNR_NN1 ._. 
</s>
<s>
Each_DD1 user_NN1 had_VHD five_MC neural_JJ networks_NN2 with_IW the_AT number_NN1 of_IO inputs_NN2 ranging_VVG from_II 2_MC to_II 6_MC ._. 
</s>
<s>
All_DB the_AT tags_NN2 mentioned_VVN in_II the_AT example_NN1 were_VBDR retained_VVN at_II the_AT coarsest_JJT granularity_NN1 level_NN1 (_( 50_MC areas_NN2 )_) ._. 
</s>
<s>
The_AT automaton_NN1 finds_VVZ the_AT occurrenceComponent=B_FO ,_, Leaf=D_FO ,_, Composite=Eafter_FO passing_VVG through_II the_AT states0,0,1,2,3_FO ,_, and4_FO ._. 
</s>
<s>
The_AT specification_NN1 uses_VVZ ontologies_NN2 and_CC an_AT1 extensible_JJ set_NN1 of_IO formalisms_NN2 to_TO present_VVI system_NN1 dynamics_NN ,_, initialize_VV0 and_CC visualize_VVI the_AT model_NN1 ._. 
</s>
<s>
Once_CS the_AT PSR-Algorithm-generated_NP1 code_NN1 has_VHZ all_DB of_IO the_AT elements_NN2 in_II a_AT1 singular_JJ vector_NN1 ,_, the_AT PSR-Algorithm_JJ inserts_NN2 code_VV0 that_CST loops_NN2 through_II the_AT vector_NN1 and_CC sets_VVZ each_DD1 SQL_NN1 input_NN1 element_NN1 to_II the_AT proper_JJ bind_NN1 variable_NN1 ._. 
</s>
<s>
The_AT LDA-based_JJ technique_NN1 correctly_RR identifiedgetTokenas_NN2 the_AT root_NN1 of_IO the_AT bug_NN1 ._. 
</s>
<s>
High_JJ level_JJ tasks_NN2 are_VBR hierarchically_RR refined_VVN into_II lower_JJR level_NN1 tasks_NN2 and_CC finally_RR into_II actions_NN2 ._. 
</s>
<s>
They_PPHS2 areproblematic_JJ to_TO specify_VVI ,_, and_CC also_RR to_TO measure_VVI ._. 
</s>
<s>
This_DD1 means_VVZ that_DD1 client_NN1 applications_NN2 use_VV0 URI_JJ identifiers_NN2 for_IF accessing_VVG the_AT resource_NN1 itself_PPX1 ._. 
</s>
<s>
An_AT1 '_NULL insider_NN1 '_NULL is_VBZ a_AT1 person_NN1 who_PNQS has_VHZ beenlegitimatelygiven_VVN the_AT capability_NN1 of_IO accessing_VVG one_MC1 or_CC many_DA2 components_NN2 of_IO the_AT IT_NN1 infrastructure_NN1 ,_, by_II interacting_VVG with_IW one_MC1 or_CC more_DAR authentication_NN1 mechanisms_NN2 ._. 
</s>
<s>
Then_RT we_PPIS2 describe_VV0 the_AT existing_JJ research_NN1 on_II detecting_VVG these_DD2 types_NN2 of_IO attacks_NN2 using_VVG collaborative_JJ intrusion_NN1 detection_NN1 systems_NN2 (_( CIDSs_NP1 )_) in_II Section3_FO ._. 
</s>
<s>
Primary_JJ data_NN collection_NN1 involved_VVD liaison_NN1 with_IW the_AT scheme_NN1 component_NN1 suppliers_NN2 ,_, and_CC included_VVD :_: product_NN1 schematics_NN2 ,_, material_NN1 quantities_NN2 ,_, energy_NN1 ratings_NN2 for_IF the_AT output_NN1 (_( message_NN1 signs_NN2 etc_RA )_) and_CC the_AT logistics_NN1 of_IO the_AT scheme_NN1 (_( considering_CS the_AT journey_NN1 time_NNT1 from_II the_AT supplier_NN1 to_II the_AT site_NN1 )_) ._. 
</s>
<s>
The_AT model_NN1 presented_VVD here_RL is_VBZ the_AT ecological_JJ component_NN1 of_IO a_AT1 wider_JJR modelling_NN1 project_NN1 that_CST aims_VVZ to_TO represent_VVI human_JJ activity_NN1 explicitly_RR in_II a_AT1 LFSM_NP1 ._. 
</s>
<s>
To_TO quantify_VVI happiness_NN1 for_IF Twitter_NN1 users_NN2 ,_, we_PPIS2 apply_VV0 the_AT real-time_JJ hedonometer_NN1 methodology_NN1 for_IF measuring_VVG sentiment_NN1 in_II large-scale_JJ text_NN1 developed_VVN in_II Dodds_NP1 et_RA21 al_RA22 ._. 
</s>
<s>
&lsqb;_( 11_MC &rsqb;_) ._. 
</s>
<s>
Initially_RR ,_, a_AT1 value_NN1 of_IO PA_NN1 and_CC BW_NP1 is_VBZ assigned_VVN to_II all_DB individuals_NN2 (_( as_CSA will_VM be_VBI detailed_VVN in_II Section_NN1 3.1_MC )_) ._. 
</s>
<s>
Finally_RR ,_, we_PPIS2 run_VV0 simulations_NN2 to_TO validate_VVI our_APPGE analytical_JJ results_NN2 for_IF less_RGR idealized_JJ networks_NN2 ._. 
</s>
<s>
The_AT filter_NN1 can_VM be_VBI passed_VVN forward_RL and_CC backward_RL over_II a_AT1 data_NN set_VV0 several_DA2 times_NNT2 and_CC the_AT number_NN1 of_IO passes_NN2 results_NN2 in_II data_NN smoothing_VVG and_CC nullification_NN1 of_IO any_DD phase_NN1 distortion_NN1 (_( Spongberg_NP1 ,_, 2000_MC )_) ._. 
</s>
<s>
Because_II21 of_II22 this_DD1 ,_, it_PPH1 was_VBDZ decided_VVN that_CST in_II this_DD1 study_NN1 WoS_NP2 would_VM also_RR be_VBI used_VVN ,_, not_XX least_RRT to_TO see_VVI if_CSW the_AT limitation_NN1 caused_VVD a_AT1 major_JJ problem_NN1 ._. 
</s>
<s>
The_AT error_NN1 estimation_NN1 becomes_VVZ more_RGR accurate_JJ as_II the_AT number_NN1 of_IO iterations_NN2 increases_NN2 and_CC it_PPH1 converges_VVZ toward_II the_AT real_JJ error_NN1 value_NN1 ._. 
</s>
<s>
The_AT comparison_NN1 showed_VVD only_RR relatively_RR minor_JJ changes_NN2 in_II the_AT resultant_JJ statistics_NN ,_, e.g._REX for_IF the_AT Viney_NP1 (_( 1991_MC )_) model_NN1 ,_, analyses_NN2 based_VVN on_II the_AT synthetic_JJ data_NN yielded_VVD a_AT1 nonlinear_JJ correlation_NN1 of_IO 0.999_MC ,_, a_AT1 mean_JJ absolute_JJ error_NN1 of_IO 0.09%_FO and_CC a_AT1 maximum_JJ absolute_JJ error_NN1 of_IO 1.02%_FO ,_, all_DB of_IO which_DDQ are_VBR similar_JJ to_II the_AT statistics_NN seen_VVN in_II Table_NN1 1_MC1 ._. 
</s>
<s>
At_II the_AT second_MD study_NN1 site_NN1 ,_, the_AT goal_NN1 was_VBDZ to_TO identify_VVI favorable_JJ drilling_NN1 areas_NN2 (_( well_RR yield_VV0 :_: >1_FO ?_? m3/h_FU ;_; low_JJ electrical_JJ conductivity_NN1 :_: <200_FO ?_? mS/m_FU )_) for_IF the_AT development_NN1 of_IO sustainable_JJ water-resources_NN2 in_II the_AT 132_MC ?_? km2_FO Juá_NP1 region_NN1 (_( Fig._NN1 ?_? 7_MC )_) ._. 
</s>
<s>
The_AT trajectories_NN2 of_IO theQ-values_NN2 obtained_VVN from_II the_AT numerical_JJ iteration_NN1 of_IO the_AT analytic_JJ model_NN1 are_VBR shown_VVN inFig._NNU 2_MC ._. 
</s>
<s>
Note_VV0 that_CST this_DD1 means_VVZ that_CST the_AT number_NN1 of_IO nodes_NN2 in_II a_AT1 cluster_NN1 is_VBZ similar_JJ to_II the_AT number_NN1 of_IO normal_JJ nodes_NN2 assigned_VVN to_II a_AT1 supernode_NN1 ._. 
</s>
<s>
Off-line_JJ processing_NN1 of_IO XML-based_JJ deployment_NN1 plans_NN2 ._. 
</s>
<s>
What_DDQ is_VBZ to_TO be_VBI monitored_VVN is_VBZ planned_VVN for_IF in_II the_AT Forensic_JJ Strategy_NN1 ,_, documented_VVN in_II the_AT Forensic_JJ Policy_NN1 ,_, and_CC prepared_VVN for_IF through_II Forensic_JJ Training_NN1 ._. 
</s>
<s>
However_RR ,_, this_DD1 does_VDZ not_XX guarantee_VVI that_CST the_AT service_NN1 will_VM validate_VVI the_AT data_NN against_II the_AT schema_NN1 at_II run-time_NN1 ._. 
</s>
<s>
Knowledge_NN1 provisioning_NN1 services_NN2 include_VV0 ontology_NN1 services_NN2 ,_, which_DDQ are_VBR in_II31 charge_II32 of_II33 the_AT storage_NN1 and_CC access_NN1 to_II the_AT conceptual_JJ models_NN2 of_IO representing_VVG knowledge_NN1 ,_, and_CC reasoning_VVG services_NN2 ,_, in_II31 charge_II32 of_II33 computational_JJ reasoning_NN1 with_IW those_DD2 conceptual_JJ models_NN2 ._. 
</s>
<s>
Nonetheless_RR ,_, to_TO broaden_VVI its_APPGE reach_NN1 requires_VVZ that_CST its_APPGE costs_NN2 be_VBI better_RRR understood_VVN and_CC where_RRQ possible_JJ reduced_JJ ._. 
</s>
<s>
They_PPHS2 disaggregate_VV0 the_AT mining_NN1 units_NN2 iteratively_RR and_CC refine_VV0 them_PPHO2 with_II31 respect_II32 to_II33 processing_NN1 ,_, up_II21 to_II22 the_AT point_NN1 that_CST the_AT refined_JJ aggregates_NN2 produce_VV0 the_AT same_DA optimal_JJ solution_NN1 for_IF the_AT linear_JJ programming_NN1 relaxation_NN1 of_IO the_AT mixed_JJ integer_NN1 programming_NN1 ._. 
</s>
<s>
Changes_NN2 of_IO interest_NN1 can_VM be_VBI highlighted_VVN by_II selecting_VVG nodes_NN2 (_( displayed_VVN in_II yellow_JJ )_) and_CC hovering_VVG the_AT mouse_NN1 over_II a_AT1 pair_NN (_( displayed_VVN in_II bold_JJ )_) ._. 
</s>
<s>
An_AT1 active_JJ defense_NN1 is_VBZ said_VVN to_II beautomatic_JJ if_CS no_AT human_JJ intervention_NN1 is_VBZ required_VVN and_CC manual_JJ if_CS key_JJ steps_NN2 require_VV0 the_AT initiation_NN1 or_CC affirmative_JJ action_NN1 of_IO humans_NN2 ._. 
</s>
<s>
This_DD1 set_NN1 includes_VVZ ontology_NN1 management_NN1 and_CC reasoning_VVG services_NN2 ,_, metadata_NN1 services_NN2 and_CC annotation_NN1 services_NN2 ._. 
</s>
<s>
In_II this_DD1 work_NN1 ,_, we_PPIS2 combine_VV0 and_CC extend_VV0 these_DD2 strategies_NN2 by_II further_RRR exploiting_VVG the_AT discrete_JJ space_NN1 splitting_NN1 of_IO the_AT underlying_JJ hierarchicalhp-finite_JJ element_NN1 basis_NN1 and_CC suggest_VV0 an_AT1 improved_JJ version_NN1 of_IO a_AT1 hierarchicalhp-finite_JJ element_NN1 solver_NN1 for_IF eddy_NN1 current_JJ problems_NN2 ._. 
</s>
<s>
This_DD1 can_VM be_VBI attributed_VVN to_II the_AT absence_NN1 of_IO queuing_VVG mechanisms_NN2 on_II the_AT return_NN1 path_NN1 from_II the_AT CCA_NP1 to_II the_AT LA_NP1 in_II31 addition_II32 to_II33 inaccuracies_NN2 in_II the_AT initial_JJ configuration_NN1 of_IO the_AT power_NN1 system_NN1 ._. 
</s>
<s>
Here_RL the_AT search_NN1 algorithm_NN1 is_VBZ a_AT1 subdivision_NN1 search_NN1 that_CST repeatedly_RR refines_VVZ the_AT domain_NN1 as_CSA indicated_VVN by_II evaluation_NN1 at_II a_AT1 grid_NN1 of_IO points_NN2 ._. 
</s>
<s>
The_AT four_MC corners_NN2 are_VBR :_: sender_NN1 ,_, sender_NN1 '_NULL s_ZZ1 AP_NP1 ,_, recipient_NN1 '_NULL s_ZZ1 AP_NP1 and_CC recipient_NN1 ._. 
</s>
<s>
To_TO accommodate_VVI all_DB the_AT values_NN2 in_II the_AT graph_NN1 ,_, a_AT1 log_NN1 scale_NN1 has_VHZ been_VBN used_VVN ._. 
</s>
<s>
Previous_JJ research_NN1 has_VHZ shown_VVN that_CST DNS_NN2 Tunnel_VV0 traffic_NN1 can_VM be_VBI differentiated_VVN from_II normal_JJ DNS_NN2 traffic_NN1 by_II analyzing_VVG character_NN1 frequencies_NN2 in_II the_AT domain_NN1 name_NN1 in_II DNS_NN2 queries_NN2 and_CC responses_NN2 (_( Born_VVN ,_, Gustafson_NP1 ,_, 2010a_FO and_CC Born_VVN ,_, Gustafson_NP1 ,_, 2010b_FO )_) ._. 
</s>
<s>
A_AT1 numerical_JJ analysis_NN1 is_VBZ also_RR presented_VVN in_II the_AT same_DA section_NN1 ._. 
</s>
<s>
Morphing_VVG chains_NN2 show_VV0 in_II detail_NN1 concept_NN1 similarity_NN1 for_IF each_DD1 aspect_NN1 ,_, and_CC how_RRQ concept_NN1 meaning_NN1 migrates_VVZ from_II one_MC1 concept_NN1 to_II another_DD1 each_DD1 year_NNT1 ,_, from_II 2003_MC to_II 2013_MC ._. 
</s>
<s>
For_REX21 instance_REX22 ,_, f12_FO will_VM increase_VVI with_IW the_AT increase_NN1 of_IO c1_FO ._. 
</s>
<s>
For_IF comparison_NN1 purposes_NN2 ,_, this_DD1 table_NN1 also_RR shows_VVZ the_AT performance_NN1 indexes_NN2 and_CC the_AT variances_NN2 of_IO the_AT residual_JJ errors_NN2 obtained_VVN from_II the_AT first_MD order_NN1 models_NN2 of_IO each_DD1 of_IO the_AT selected_JJ model_NN1 structures_NN2 j_ZZ1 ._. 
</s>
<s>
Table_NN1 ?_? 1_MC1 shows_VVZ that_CST the_AT best_JJT performance_NN1 index_NN1 (_( 87.08%_FO )_) was_VBDZ obtained_VVN by_II the_AT second_MD order_NN1 nominal_JJ model_NN1 with_IW the_AT ARMAX_NN1 structure_NN1 ._. 
</s>
<s>
Whereas_CS this_DD1 setup_NN1 constitutes_VVZ a_AT1 well-constrained_JJ model_NN1 selection_NN1 problem_NN1 due_II21 to_II22 the_AT large_JJ number_NN1 of_IO observations_NN2 relative_II21 to_II22 covariates_NN2 ,_, the_AT key_NN1 here_RL is_VBZ that_CST such_DA a_AT1 setup_NN1 would_VM not_XX be_VBI computationally_RR feasible_JJ if_CS one_PN1 were_VBDR to_TO examine_VVI all_DB possible_JJ models_NN2 ._. 
</s>
<s>
Since_CS we_PPIS2 believe_VV0 that_DD1 motion_NN1 can_VM be_VBI recognized_VVN directly_RR without_IW knowing_VVG a_JJ21 priori_JJ22 human_JJ geometric_JJ information_NN1 ,_, our_APPGE focus_NN1 is_VBZ on_II choosing_VVG features_NN2 that_CST can_VM describe_VVI the_AT dynamic_JJ properties_NN2 of_IO motion_NN1 to_TO build_VVI the_AT model_NN1 ._. 
</s>
<s>
In_II a_AT1 two_MC dimensional_JJ domain_NN1 of_IO dimensionL×H_FO ,_, the_AT constant_JJ average_JJ particle_NN1 volumeV0is_FO given_VVN byV0=LH/N_FU ,_, i.e._REX the_AT total_JJ domain_NN1 area_NN1 is_VBZ distributed_VVN equally_RR to_II each_DD1 particle_NN1 ._. 
</s>
<s>
We_PPIS2 have_VH0 presented_VVN a_AT1 novel_JJ multigranularity_NN1 locking_VVG model_NN1 suitable_JJ for_IF applications_NN2 that_CST manipulate_VV0 RDF_NP1 data_NN ._. 
</s>
<s>
All_DB important_JJ constructs_NN2 and_CC their_APPGE hypothesized_JJ relationships_NN2 are_VBR shown_VVN schematically_RR in_II Fig._NN1 2_MC ._. 
</s>
<s>
We_PPIS2 define_VV0 the_AT vocabulary_NN1 W_ZZ1 as_II the_AT list_NN1 of_IO the_AT symbols_NN2 ,_, or_CC words_NN2 ,_, that_DD1 can_VM be_VBI in_II the_AT sentence_NN1 ._. 
</s>
<s>
All_DB processes_NN2 only_RR affect_VV0 the_AT non-ionized_JJ fraction_NN1 of_IO MPs_NNU ._. 
</s>
<s>
NPV_NP1 has_VHZ been_VBN used_VVN ,_, among_II other_JJ performance_NN1 criteria_NN2 ,_, to_TO measure_VVI the_AT performance_NN1 of_IO the_AT model_NN1 ._. 
</s>
<s>
In_RR21 particular_RR22 the_AT use_NN1 of_IO advanced_JJ spatial_JJ burn_NN1 probability_NN1 modeling_NN1 techniques_NN2 is_VBZ gaining_VVG popularity_NN1 (_( Carmel_NP1 et_RA21 al._RA22 ,_, 2009_MC ,_, Scott_NP1 et_RA21 al._RA22 ,_, 2012a_FO ,_, Parisien_NP1 et_RA21 al._RA22 ,_, 2013_MC )_) ._. 
</s>
<s>
In_II the_AT case_NN1 of_IO non-live_JJ processing_NN1 ,_, the_AT collected_JJ JSON_NN1 is_VBZ processed_VVN using_VVG the_AT GATE_NN1 Cloud_NN1 Paralleliser_NN1 (_( GCP_NP1 )_) to_TO load_VVI the_AT JSON_NN1 files_NN2 into_II GATE_NN1 documents_NN2 (_( one_MC1 document_NN1 per_II tweet_NN1 )_) ,_, annotate_VV0 them_PPHO2 ,_, and_CC then_RT index_VV0 them_PPHO2 for_IF search_NN1 and_CC visualisation_NN1 in_II the_AT GATE_NN1 Mímir_NP1 framework_NN1 &lsqb;_( 13_MC &rsqb;_) ._. 
</s>
<s>
This_DD1 procedure_NN1 is_VBZ repeated_VVN until_CS the_AT work_NN1 amount_NN1 of_IO the_AT household_NN1 has_VHZ reached_VVN 1_MC1 or_CC the_AT household_NN1 has_VHZ gone_VVN through_II all_DB observed_JJ firms_NN2 ._. 
</s>
<s>
Hence_RR ,_, we_PPIS2 think_VV0 that_CST the_AT idea_NN1 of_IO the_AT accelerator_NN1 (_( positive_JJ approximation_NN1 )_) is_VBZ promising_JJ to_TO be_VBI applied_VVN to_II other_JJ types_NN2 of_IO forward_JJ search_NN1 feature_NN1 selection_NN1 approaches_NN2 based_VVN on_II measures_NN2 like_II information_NN1 gain_NN1 ,_, distance_NN1 ,_, dependency_NN1 and_CC information_NN1 entropy_NN1 ._. 
</s>
<s>
Our_APPGE projects_NN2 also_RR contained_VVN vulnerability_NN1 types_NN2 that_CST are_VBR considered_VVN design_NN1 flaws_NN2 &lsqb;_( 10_MC &rsqb;_) ._. 
</s>
<s>
Our_APPGE proposal_NN1 produces_VVZ great_JJ results_NN2 in_II this_DD1 dataset_NN1 ,_, surpassing_VVG the_AT rest_NN1 of_IO the_AT methods_NN2 in_II every_AT1 pairing_NN1 except_II21 for_II22 the_AT pathological_JJ '_NULL Yes_UH '_NULL vs_II '_NULL No_AT '_NULL case_NN1 ._. 
</s>
<s>
PRNG(EPCsNT)PX_VV0 ?_? and_CC ?_? 
</s>
<s>
Info=_FO (_( DATARID_NP1 )_) ?_? and_CC forwards_VVZ them_PPHO2 to_II the_AT reader_NN1 ._. 
</s>
<s>
The_AT models_NN2 are_VBR built_VVN incrementally_RR as_CSA each_DD1 new_JJ data_NN instance_NN1 (_( software_NN1 build_NN1 )_) that_CST arrives_VVZ is_VBZ used_VVN to_TO update_VVI the_AT existing_JJ model_NN1 ._. 
</s>
<s>
We_PPIS2 use_VV0 the_AT algorithms_NN2 given_VVN in_II Section_NN1 4_MC to_TO get_VVI upper_JJ bounds_NN2 ._. 
</s>
<s>
The_AT same_DA set_NN1 Ei_NP1 of_IO edges_NN2 is_VBZ entirely_RR visible_JJ from_II each_DD1 point_NN1 of_IO Zi_NP1 ,_, ?_? i_ZZ1 ._. 
</s>
<s>
Daniel_NP1 Aceitunaa_NP1 ,_, Gursimran_NP1 Waliaa_NP1 ,_, Hyunsook_NP1 Doa_NP1 ,_, Seok-Won_JJ Leeb_NP1 ,_, We_PPIS2 wanted_VVD subjects_NN2 to_TO know_VVI what_DDQ surface_NN1 they_PPHS2 were_VBDR navigating_VVG so_RG a_AT1 2D_NNU shape_NN1 ,_, representing_VVG the_AT current_JJ surface_NN1 ,_, was_VBDZ placed_VVN around_II the_AT north_ND1 pointing_JJ arrow_NN1 (_( see_VV0 letter_NN1 A_ZZ1 in_II @a_FO and_CC d_ZZ1 )_) ._. 
</s>
<s>
If_CS the_AT count_NN1 of_IO any_DD arc_NN1 becomes_VVZ equal_JJ to_TO zero_VVI ,_, then_RT that_DD1 arc_NN1 should_VM be_VBI removed_VVN from_II the_AT set_NN1 of_IO arcs_NN2 associated_VVN with_IW the_AT corresponding_JJ session_NNT1 ._. 
</s>
<s>
We_PPIS2 believe_VV0 the_AT prediction_NN1 accuracy_NN1 could_VM be_VBI improved_VVN by_II obtaining_VVG more_RRR and_CC better_JJR quality_NN1 data_NN and_CC by_II exploring_VVG more_DAR categories_NN2 of_IO metrics_NN2 ._. 
</s>
<s>
True_JJ to_II the_AT Dartmouth_NP1 vision_NN1 of_IO simulation_NN1 ,_, college_NN1 sophomores_NN2 were_VBDR asked_VVN to_TO think_VVI aloud_RR as_CSA they_PPHS2 solved_VVD symbolic_JJ logic_NN1 problems_NN2 ,_, and_CC GPS_NN2 was_VBDZ developed_VVN to_TO simulate_VVI what_DDQ Newell_NP1 and_CC Simon_NP1 observed_VVD ._. 
</s>
<s>
The_AT main_JJ limitation_NN1 of_IO the_AT proposal_NN1 is_VBZ thus_RR that_CST even_CS21 when_CS22 the_AT models_NN2 serve_VV0 as_CSA information_NN1 repositories_NN2 ,_, there_EX is_VBZ no_AT conclusion_NN1 of_IO the_AT relations_NN2 expressed_VVN on_II them_PPHO2 ._. 
</s>
<s>
However_RR ,_, as_CSA can_VM also_RR be_VBI seen_VVN by_II some_DD of_IO our_APPGE properties_NN2 ,_, it_PPH1 is_VBZ often_RR difficult_JJ to_TO decide_VVI whether_CSW31 or_CSW32 not_CSW33 a_AT1 formula_NN1 is_VBZ a_AT1 safety_NN1 property_NN1 ,_, whether_CSW it_PPH1 is_VBZ co-safety_NN1 ,_, or_CC neithereven_NN1 for_IF the_AT trained_JJ eye_NN1 ._. 
</s>
<s>
The_AT goal_NN1 of_IO Labeled-SSiPP_NP1 is_VBZ to_TO improve_VVI the_AT convergence_NN1 time_NNT1 of_IO SSiPP_NN1 to_II the_AT -consistent_JJ solution_NN1 ._. 
</s>
<s>
These_DD2 rules_NN2 can_VM be_VBI imposed_VVN from_II the_AT sole_JJ knowledge_NN1 of_IO the_AT route_NN1 ,_, the_AT RTT_NP1 of_IO each_DD1 flow_NN1 ,_, the_AT characteristics_NN2 of_IO each_DD1 router_NN1 ,_, and_CC the_AT link_NN1 characteristics_NN2 (_( buffer_NN1 size_NN1 ,_, link_VV0 capacity_NN1 ,_, scheduling_VVG ,_, etc._RA )_) in_II the_AT network_NN1 ._. 
</s>
<s>
There_EX are_VBR potentially_RR doubly-exponentially_RR many_DA2 states_NN2 in_II the_AT size_NN1 of_IO the_AT domain_NN1 ?_? &lsqb;_( 12_MC &rsqb;_) ,_, meaning_VVG possibly_RR doubly-exponential_JJ increase_NN1 in_II the_AT size_NN1 of_IO the_AT determinization_NN1 of_IO the_AT nondeterministic_JJ planning_NN1 domain_NN1 ._. 
</s>
<s>
Each_DD1 timeslot_NN1 contains_VVZ one_PN1 more_RGR injected_JJ anomaly_NN1 than_CSN the_AT previous_JJ one_PN1 (_( see_VV0 also_RR Section_NN1 5_MC )_) ._. 
</s>
<s>
Both_DB2 the_AT cut_NN1 enumeration_NN1 technique_NN1 and_CC genetic_JJ algorithm_NN1 provide_VV0 a_AT1 higher_JJR lower_JJR bound_NN1 than_CSN the_AT Monte_NP1 Carlo_NP1 simulation_NN1 ,_, with_IW the_AT genetic_JJ algorithm_NN1 being_VBG the_AT higher_JJR of_IO the_AT two_MC ._. 
</s>
<s>
In_II this_DD1 context_NN1 ,_, a_AT1 planning_NN1 problem_NN1 world_NN1 state_NN1 can_VM be_VBI described_VVN using_VVG an_AT1 interpretation_NN1 providing_VVG the_AT objects_NN2 in_II the_AT world_NN1 and_CC the_AT relations_NN2 among_II them_PPHO2 ._. 
</s>
<s>
First_MD ,_, one_MC1 author_NN1 '_NULL s_ZZ1 personal_JJ experience_NN1 was_VBDZ that_CST paired_VVD programming_NN1 is_VBZ valuable_JJ for_IF the_AT development_NN1 of_IO complex_JJ software_NN1 while_CS the_AT case_NN1 study_NN1 from_II another_DD1 author_NN1 showed_VVD that_CST it_PPH1 is_VBZ not_XX useful_JJ for_IF scientific_JJ software_NN1 development_NN1 ._. 
</s>
<s>
The_AT execution_NN1 of_IO some_DD active_JJ defenses_NN2 may_VM even_RR be_VBI restricted_VVN to_II government_NN1 agencies_NN2 with_IW appropriate_JJ authorities_NN2 and_CC responsibilities_NN2 ._. 
</s>
<s>
We_PPIS2 will_VM refer_VVI to_II the_AT minimal_JJ travel_NN1 time_NNT1 from_II Li_NP1 to_II Lj_NP1 as_CSA Tij_NP1 ._. 
</s>
<s>
With_IW CMAES_NP2 ,_, some_DD of_IO the_AT global_JJ optimization_NN1 trials_NN2 were_VBDR prematurely_RR terminated_VVN ,_, but_CCB only_RR after_II the_AT optimization_NN1 process_NN1 had_VHD progressed_VVN to_II a_AT1 point_NN1 wherein_RRQ the_AT known_JJ global_JJ min_NNU ._. 
</s>
<s>
Table_NN1 8_MC presents_VVZ the_AT comparative_JJ study_NN1 ._. 
</s>
<s>
To_TO simplify_VVI notation_NN1 ,_, we_PPIS2 use_VV0 the_AT letter_NN1 "_" p_ZZ1 "_" (_( as_CSA in_II "_" planned_JJ "_" )_) in_II31 place_II32 of_II33 the_AT letter_NN1 "_" i_ZZ1 "_" to_TO represent_VVI planned_VVN (_( intention_NN1 )_) locks_NN2 ._. 
</s>
<s>
This_DD1 is_VBZ illustrated_VVN in_II the_AT plot_NN1 in_II (_( a_ZZ1 )_) ,_, which_DDQ shows_VVZ a_AT1 reflected_JJ wave_NN1 behind_II the_AT expansion_NN1 wavefront_NN1 at_II the_AT bottom_JJ boundary_NN1 ._. 
</s>
<s>
This_DD1 corresponds_VVZ to_II the_AT Young_NP1 '_NULL s_ZZ1 modulus_NN1 of_IO Y=30=15–36N/m_FU due_II21 to_II22 the_AT use_NN1 of_IO a_AT1 three-dimensional_JJ membrane_NN1 model_NN1 ._. 
</s>
<s>
This_DD1 approach_NN1 distinguishes_VVZ datalegend_NN1 from_II other_JJ data_NN integration_NN1 efforts_NN2 :_: individual_JJ users_NN2 can_VM contribute_VVI their_APPGE own_DA links_NN2 through_II the_AT system_NN1 ,_, leading_VVG to_II a_AT1 growing_JJ ,_, heterogeneous_JJ but_CCB interconnected_VVD web_NN1 of_IO linked_JJ statistical_JJ data_NN ._. 
</s>
<s>
It_PPH1 is_VBZ calculated_VVN by_II determining_VVG ,_, for_IF each_DD1 taxonomic_JJ group_NN1 ,_, the_AT proportion_NN1 of_IO individuals_NN2 that_CST it_PPH1 contributes_VVZ to_II the_AT total_JJ sample_NN1 ._. 
</s>
<s>
In_RR21 addition_RR22 ,_, an_AT1 AHE_NN1 web_NN1 client_NN1 has_VHZ been_VBN developed_VVN to_TO provide_VVI a_AT1 simple_JJ interface_NN1 for_IF the_AT end_NN1 user_NN1 when_CS interacting_VVG with_IW the_AT AHE_NN1 server_NN1 via_II a_AT1 web_NN1 browser_NN1 ._. 
</s>
<s>
In_II our_APPGE evaluation_NN1 ,_, we_PPIS2 assume_VV0 that_CST the_AT maximum_JJ allowable_JJ transaction_NN1 completion_NN1 times_NNT2 are_VBR the_AT SLA_NN1 constraints_NN2 on_II the_AT resource_NN1 allocation_NN1 problem_NN1 ._. 
</s>
<s>
The_AT authors_NN2 use_VV0 the_AT network_NN1 '_NULL s_ZZ1 context_NN1 (_( node_NN1 number_NN1 and_CC mobility_NN1 )_) as_CSA inputs_NN2 in_II31 conjunction_II32 with_II33 the_AT network_NN1 '_NULL s_ZZ1 performance_NN1 (_( delay_VV0 ,_, routing_VVG delivery_NN1 rate_NN1 ,_, routing_VVG packets_NN2 delivery_NN1 rate_NN1 and_CC routing_VVG load_NN1 )_) as_CSA outputs_VVZ for_IF modelling_NN1 ._. 
</s>
<s>
This_DD1 will_NN1 ,_, in_II turn_NN1 ,_, impact_NN1 how_RRQ solids_NN2 and_CC fluids_NN2 interact_VV0 with_IW the_AT tissue_NN1 barriers_NN2 ._. 
</s>
<s>
Lets_VVZ denote_VVI ith_MD block_NN1 of_IO the_AT vector_NN1 by_II Vi_NP1 ._. 
</s>
<s>
This_DD1 random_JJ behavior_NN1 is_VBZ because_CS the_AT network_NN1 is_VBZ a_AT1 random_JJ network_NN1 and_CC deleting_VVG some_DD nodes_NN2 through_II RKD_NP1 might_VM generate_VVI different_JJ structure_NN1 for_IF the_AT resulting_JJ network_NN1 ,_, and_CC therefore_RR ,_, resulting_VVG different_JJ values_NN2 for_IF after_II deleting_VVG each_DD1 node_NN1 ._. 
</s>
<s>
We_PPIS2 also_RR compare_VV0 our_APPGE approach_NN1 with_IW a_AT1 modern_JJ reactive_JJ control_NN1 strategy_NN1 that_CST is_VBZ used_VVN in_II industrial_JJ power_NN1 systems_NN2 management_NN1 ._. 
</s>
<s>
Integratedapproaches_NN2 wherein_RRQ the_AT programmer_NN1 builds_VVZ understanding_VVG by_II switching_VVG between_II top_NN1 down_RP ,_, bottom_VV0 up_RP and_CC knowledge_NN1 based_VVN approaches&lsqb;44&rsqb;_FO ;_; or_CC in_II which_DDQ chunking_NN1 and_CC tracing_VVG processes_NN2 are_VBR intertwined_VVN :_: chunking_VVG for_IF comprehension_NN1 and_CC tracing_VVG to_TO identify_VVI related_JJ chunks_NN2 ,_, ripple_NN1 effects_NN2 from_II any_DD modifications_NN2 ,_, and_CC faults&lsqb;4&rsqb;_FO ._. 
</s>
<s>
Throughout_II the_AT simultaneous_JJ search_NN1 procedure_NN1 ,_, all_DB accepted_JJ transitions_NN2 from_II the_AT native_JJ state_NN1 to_II the_AT current_JJ conformation_NN1 are_VBR recorded_VVN for_IF each_DD1 individual_NN1 ._. 
</s>
<s>
This_DD1 characteristic_NN1 means_VVZ that_CST a_AT1 mobile_JJ device_NN1 enjoys_VVZ a_AT1 longer_JJR term_NN1 ,_, persistent_JJ relationship_NN1 with_IW a_AT1 user_NN1 that_CST can_VM be_VBI exploited_VVN for_IF authentication_NN1 purposes_NN2 ._. 
</s>
<s>
One_MC1 of_IO the_AT earliest_JJT such_DA works_NN is_VBZ &lsqb;_( 32_MC &rsqb;_) ._. 
</s>
<s>
Profiles_NN2 store_VV0 the_AT parameters_NN2 according_II21 to_II22 which_DDQ images_NN2 are_VBR processed_VVN ._. 
</s>
<s>
The_AT typical_JJ SSL_NP1 use_NN1 includes_VVZ server_NN1 authentication_NN1 ;_; newer_JJR SSL_NP1 uses_NN2 permit_VV0 the_AT browser_NN1 to_TO authenticate_VVI via_II PKI_NP1 as_RR21 well_RR22 ._. 
</s>
<s>
Results_NN2 of_IO this_DD1 pilot_NN1 indicated_VVD that_CST the_AT public_JJ sector_NN1 had_VHD not_XX yet_RR tapped_VVN into_II the_AT full_JJ potential_NN1 of_IO its_APPGE address_NN1 registries_NN2 ._. 
</s>
<s>
It_PPH1 is_VBZ also_RR called_VVN Social_JJ Contagion_NN1 &lsqb;_( 15_MC &rsqb;_) ._. 
</s>
<s>
That_REX21 is_REX22 ,_, the_AT 75/425_MF split_NN1 is_VBZ not_XX to_TO be_VBI intended_VVN as_CSA training_NN1 set_VVN versus_II validation_NN1 set_NN1 ,_, but_CCB as_CSA the_AT extension_NN1 of_IO the_AT training_NN1 set_VVN versus_II validation_NN1 set_NN1 ._. 
</s>
<s>
Tahmasebi_NN2 and_CC Hezarkhani_NP1 (_( 2010_MC )_) suggest_VV0 the_AT possibility_NN1 of_IO overcoming_VVG these_DD2 problems_NN2 by_II optimizing_VVG artificial_JJ neural_JJ network_NN1 parameters_NN2 using_VVG a_AT1 genetic_JJ algorithm_NN1 ._. 
</s>
<s>
This_DD1 is_VBZ so_RR since_CS the_AT terms_NN2 x_ZZ1 and_CC y_ZZ1 have_VH0 a_AT1 polar_JJ dependence_NN1 and_CC hence_RR cause_VV0 a_AT1 milder_JJR singularity_NN1 ._. 
</s>
<s>
In_II others_NN2 ,_, search_NN1 is_VBZ used_VVN to_TO figure_VVI out_RP the_AT best_JJT routine_NN1 &lsqb;_( 11_MC &rsqb;_) ._. 
</s>
<s>
Following_VVG successful_JJ approval_NN1 the_AT completion_NN1 of_IO registration_NN1 Identifier_NN1 Credential_JJ issuing_NN1 ,_, re-activation_NN1 and_CC revocation_NN1 are_VBR essential_JJ activities_NN2 of_IO APIM_NP1 management_NN1 ._. 
</s>
<s>
Dou_NP1 Di_NP1 Zhu_NP1 is_VBZ a_AT1 three-player_JJ ladder-based_NN1 (_( orclimbing_VVG )_) card_NN1 game_NN1 ,_, played_VVN with_IW a_AT1 standard_JJ deck_NN1 of_IO 52_MC cards_NN2 plus_II two_MC jokers_NN2 ,_, hugely_RR popular_JJ in_II ._. 
</s>
<s>
Our_APPGE results_NN2 show_VV0 that_CST this_DD1 version_NN1 of_IO the_AT model_NN1 is_VBZ functioning_VVG as_CSA intended_VVN and_CC have_VH0 highlighted_VVN which_DDQ parameters_NN2 have_VH0 greatest_JJT influence_NN1 on_II two_MC aspects_NN2 of_IO the_AT model_NN1 ;_; land-cover_JJ change_NN1 and_CC the_AT wildfire_NN1 regime_NN1 ._. 
</s>
<s>
In_II parallel_NN1 ,_, the_AT collector_NN1 writes_VVZ the_AT tweet_NN1 JSON_NN1 to_II a_AT1 backup_NN1 file_NN1 ,_, so_CS it_PPH1 is_VBZ preserved_VVN for_IF future_JJ reference_NN1 (_( for_REX21 example_REX22 ,_, if_CS we_PPIS2 improve_VV0 the_AT analysis_NN1 pipeline_NN1 we_PPIS2 may_VM want_VVI to_TO go_VVI back_RP and_CC re-process_VVI previously-collected_JJ tweets_NN2 with_IW the_AT new_JJ pipeline_NN1 )_) ._. 
</s>
<s>
Word_NN1 clouds_NN2 have_VH0 been_VBN used_VVN to_TO assist_VVI users_NN2 in_II browsing_VVG social_JJ media_NN streams_NN2 ,_, including_II blog_NN1 content_NN1 &lsqb;_( 44_MC &rsqb;_) and_CC tweets_NN2 &lsqb;_( 45_MC &rsqb;_) ,_, &lsqb;_( 46_MC &rsqb;_) ._. 
</s>
<s>
For_IF a_AT1 finite_JJ prefix_NN1 w∈_FO (_( S_ZZ1 A_ZZ1 )_) S_ZZ1 of_IO a_AT1 play_NN1 ,_, we_PPIS2 denote_VV0 by_II Cone(w)_NP1 the_AT set_NN1 of_IO plays_NN2 with_IW w_ZZ1 as_II the_AT prefix_NN1 (_( i.e._REX ,_, the_AT cone_NN1 or_CC cylinder_NN1 of_IO the_AT prefix_NN1 w_ZZ1 )_) ,_, and_CC denote_VV0 by_II Last(w)_NP1 the_AT last_MD state_NN1 of_IO w_ZZ1 ._. 
</s>
<s>
A_AT1 function_NN1 of_IO number_NN1 of_IO hops_NN2 against_II number_NN1 of_IO aggregation_NN1 points_NN2 is_VBZ given_VVN in_II Fig._NN1 3_MC ._. 
</s>
<s>
In_II the_AT case_NN1 of_IO multiple_JJ neighborhoods_NN2 ,_, a_AT1 data_NN fusion_NN1 approach_NN1 is_VBZ adopted_VVN to_TO combine_VVI the_AT results_NN2 of_IO more_DAR than_CSN one_MC1 neighborhoods_VVZ ._. 
</s>
<s>
In_II Section_NN1 2_MC we_PPIS2 survey_VV0 the_AT most_RGT important_JJ related_JJ work_NN1 to_II the_AT five_MC core_NN1 aspects_NN2 addressed_VVN by_II our_APPGE platform_NN1 ._. 
</s>
<s>
In_II a_AT1 recent_JJ paper_NN1 Harpaz_NP1 and_CC Haralick_NP1 &lsqb;_( 12_MC &rsqb;_) demonstrated_VVD how_RRQ the_AT shifting_NN1 and_CC amplification_NN1 patterns_NN2 along_II21 with_II22 patterns_NN2 that_CST induce_VV0 negative_JJ or_CC more_RGR complex_JJ correlations_NN2 can_VM be_VBI generalized_VVN to_II linear_JJ manifolds_NN2 ,_, making_VVG the_AT paradigm_NN1 of_IO linear_JJ manifold_JJ clustering_NN1 applicable_JJ also_RR to_II the_AT problem_NN1 of_IO pattern/correlation_NN1 clustering_NN1 ._. 
</s>
<s>
In_II31 terms_II32 of_II33 reliability_NN1 ,_, we_PPIS2 have_VH0 recognised_VVN in_II Section_NN1 3.5_MC that_CST we_PPIS2 can_VM not_XX describe_VVI in_II detail_NN1 the_AT data_NN collection_NN1 and_CC analysis_NN1 activities_NN2 we_PPIS2 went_VVD through_RP ._. 
</s>
<s>
More_RGR recently_RR ,_, linguistic_JJ terms_NN2 such_II21 as_II22 '_NULL high_JJ '_NULL or_CC '_NULL very_RG high_RR '_NULL came_VVD into_II use_NN1 ,_, and_CC we_PPIS2 make_VV0 use_NN1 of_IO them_PPHO2 in_II this_DD1 paper_NN1 ._. 
</s>
<s>
Stakeholder_VV0 participants_NN2 were_VBDR involved_JJ in_II the_AT decision_NN1 making_VVG process_NN1 at_II various_JJ stages_NN2 ,_, including_II model_NN1 selection_NN1 and_CC development_NN1 ,_, data_NN collection_NN1 and_CC integration_NN1 ,_, scenario_NN1 development_NN1 ,_, interpretation_NN1 of_IO results_NN2 ,_, and_CC development_NN1 of_IO policy_NN1 alternatives_NN2 ._. 
</s>
<s>
Agents_NN2 equipped_VVN with_IW BBNs_NP1 '_NULL brains_NN2 '_NULL make_VV0 decisions_NN2 using_VVG both_RR qualitative_JJ empirical_JJ information_NN1 ,_, e.g._REX beliefs_NN2 and_CC attitudes_NN2 of_IO stakeholders_NN2 derived_VVN from_II participatory_JJ workshops_NN2 ,_, and_CC quantitative_JJ data_NN collected_VVN via_II household_NN1 surveys_NN2 ._. 
</s>
<s>
In_II this_DD1 technique_NN1 ,_, MSNitransmits_NN2 aDiscovermessage_VV0 with_IW atime-to-live(TTL)_JJ field_NN1 to_II its_APPGE parent_NN1 on_II the_AT tree_NN1 ._. 
</s>
<s>
The_AT methodologies_NN2 discussed_VVN in_II the_AT preceding_JJ sections_NN2 were_VBDR further_RRR demonstrated_VVN by_II applying_VVG them_PPHO2 to_II the_AT calibration_NN1 of_IO a_AT1 GSSHA_NN1 hydrologic_JJ model_NN1 for_IF the_AT Goodwin_NP1 Creek_NN1 Experimental_JJ Watershed_NN1 (_( GCEW_NP1 )_) (_( &lsqb;_( Senarath_NP1 et@al._FO ,_, 2000_MC &rsqb;_) and_CC &lsqb;_( Downer_NP1 and_CC Ogden_NP1 ,_, 2003b_FO &rsqb;_) )_) ._. 
</s>
<s>
The_AT same_DA tolerances_NN2 are_VBR used_VVN for_IF the_AT tangent_JJ linear_JJ and_CC adjoint_JJ coarse_JJ and_CC fine_JJ solvers_NN2 ,_, respectively_RR ._. 
</s>
<s>
Models_NN2 such_II21 as_II22 DAG_NP1 models_NN2 represent_VV0 a_AT1 middle_JJ ground_NN1 between_II the_AT two_MC approaches_NN2 ._. 
</s>
<s>
Characterization_NN1 of_IO existing_JJ empirical_JJ research_NN1 ,_, evaluating_VVG the_AT use_NN1 of_IO software_NN1 patterns_NN2 by_II human_JJ participants_NN2 ,_, is_VBZ a_AT1 first_MD step_NN1 towards_II understanding_VVG the_AT dynamics_NN of_IO pattern_NN1 application_NN1 in_II problem_NN1 solving_NN1 ._. 
</s>
<s>
Although_CS some_DD roughness_NN1 values_VVZ greater_JJR than_CSN 1.0_MC are_VBR found_VVN for_IF some_DD input_NN1 values_NN2 (_( e.g._REX Fr_NPD1 ,_, highlighted_VVN in_II Table_NN1 ?_? 2_MC )_) ,_, indicating_VVG relationships_NN2 between_II inputs_NN2 and_CC outputs_NN2 ,_, none_PN are_VBR large_JJ enough_RR to_TO indicate_VVI extreme_JJ non-linearity_NN1 ._. 
</s>
<s>
The_AT authors_NN2 showed_VVD that_CST considering_CS effort_NN1 yields_VVZ different_JJ results_NN2 from_II standard_JJ bug_NN1 prediction_NN1 (_( i.e._REX ,_, the_AT performance_NN1 of_IO models_NN2 built_VVD using_VVG size_NN1 metrics_NN1 becomes_VVZ comparable_JJ to_II that_DD1 of_IO a_AT1 random_JJ predictor_NN1 )_) ._. 
</s>
<s>
The_AT first_MD is_VBZ coarse-grained_JJ parallelism_NN1 ,_, which_DDQ uses_VVZ a_AT1 single_JJ parallel_JJ section_NN1 covering_VVG the_AT entire_JJ execution_NN1 of_IO f_ZZ1 ,_, as_CSA was_VBDZ done_VDN in_II ,_, for_REX21 example_REX22 &lsqb;_( 11_MC &rsqb;_) ._. 
</s>
<s>
These_DD2 apprehensions_NN2 can_VM hinder_VVI participants_NN2 '_NULL potential_JJ to_TO perform_VVI to_II the_AT best_JJT of_IO their_APPGE abilities_NN2 whether_CSW they_PPHS2 are_VBR based_VVN on_II actual_JJ or_CC perceived_VVD liabilities_NN2 associated_VVN with_IW participation_NN1 in_II the_AT experiment_NN1 ._. 
</s>
<s>
We_PPIS2 now_RT discuss_VV0 the_AT actions_NN2 executed_VVN by_II the_AT AP_NP1 under_II the_AT two_MC different_JJ scenarios_NN2 ._. 
</s>
<s>
One_MC1 rather_RG serious_JJ problem_NN1 is_VBZ that_CST ,_, unlike_JJ OWL_NN1 ,_, rules_NN2 have_VH0 variables_NN2 ,_, so_RR treating_VVG them_PPHO2 as_II a_AT1 semantic_JJ extension_NN1 of_IO RDF_NP1 is_VBZ very_RG difficult_JJ ._. 
</s>
<s>
Our_APPGE findings_NN2 reveal_VV0 that_CST there_EX is_VBZ a_AT1 continuum_NN1 of_IO levels_NN2 of_IO customer_NN1 involvement_NN1 on_II real-life_JJ Agile_JJ projects_NN2 (_( Fig._NN1 5_MC )_) ._. 
</s>
<s>
To_TO improve_VVI the_AT performance_NN1 the_AT idea_NN1 is_VBZ to_TO re-use_VVI the_AT result_NN1 of_IO the_AT SPRT_NP1 that_CST has_VHZ been_VBN obtain_VV0 after_II the_AT last_MD monitoring_NN1 run_NN1 ._. 
</s>
<s>
For_IF any_DD node_NN1 ,_, the_AT closest_JJT neighbour_NN1 on_II the_AT tree_NN1 (_( minimum_JJ tree_NN1 distance_NN1 )_) is_VBZ its_APPGE primary_JJ parent_NN1 node_NN1 or_CC its_APPGE immediate_JJ child_NN1 nodes_NN2 ,_, all_RR having_VHG a_AT1 tree_NN1 distance_NN1 of_IO one_MC1 hop_NN1 from_II this_DD1 node_NN1 ._. 
</s>
<s>
To_II the_AT extent_NN1 that_CST the_AT issue_NN1 at_II hand_NN1 is_VBZ ,_, for_REX21 example_REX22 ,_, the_AT extraction_NN1 of_IO content_NN1 from_II a_AT1 different_JJ and_CC unrelated_JJ database_NN1 on_II the_AT machine_NN1 containing_VVG the_AT SMTP_NP1 server_NN1 ,_, this_DD1 is_VBZ not_XX within_II the_AT syntax_NN1 specifiable_JJ by_II the_AT SMTP_NP1 protocol_NN1 or_CC the_AT envelope_NN1 of_IO control_NN1 of_IO a_AT1 typical_JJ SMTP_NP1 server_NN1 from_II the_AT user_NN1 (_( in_II this_DD1 case_NN1 external_JJ server_NN1 )_) interface_VV0 ._. 
</s>
<s>
Having_VHG said_VVN this_DD1 ,_, however_RR ,_, it_PPH1 is_VBZ also_RR worth_II noting_VVG approximately_RR two_MC thirds_MF of_IO those_DD2 that_CST had_VHD heard_VVN of_IO the_AT Get_VV0 Safe_JJ Online_JJ site_NN1 classed_VVN themselves_PPX2 as_CSA '_NULL advanced_JJ '_NULL users_NN2 ,_, suggesting_VVG that_CST the_AT users_NN2 most_RGT likely_JJ to_TO be_VBI in_II31 need_II32 of_II33 assistance_NN1 may_VM be_VBI failing_VVG to_TO receive_VVI the_AT message_NN1 ._. 
</s>
<s>
By_II comparing_VVG one_MC1 technique_NN1 against_II another_DD1 ,_, specific_JJ comments_NN2 on_II strengths_NN2 and_CC weaknesses_NN2 could_VM be_VBI made_VVN ,_, as_II31 well_II32 as_II33 conclusions_NN2 regarding_II the_AT best_JJT approach_NN1 to_II security_NN1 testing_NN1 ._. 
</s>
<s>
The_AT gPC_NNU method_NN1 has_VHZ been_VBN successfully_RR implemented_VVN to_TO analyze_VVI the_AT uncertainty_NN1 quantification_NN1 for_IF problems_NN2 such_II21 as_II22 fluid_NN1 flows_VVZ with_IW uncertain_JJ initial_NN1 and_CC boundary_NN1 conditions_NN2 ,_, sensitivity_NN1 analysis_NN1 and_CC steady-state_JJ problems_NN2 &lsqb;_( 6_MC &rsqb;_) ,_, &lsqb;_( 44_MC &rsqb;_) ,_, &lsqb;_( 46_MC &rsqb;_) ,_, &lsqb;_( 47_MC &rsqb;_) and_CC &lsqb;_( 48_MC &rsqb;_) ._. 
</s>
<s>
All_DB pseudoinvariants_NN2 of_IO any_DD object_NN1 with_IW at_RR21 least_RR22 one_MC1 axis_NN1 of_IO symmetry_NN1 are_VBR identically_RR zero_MC ._. 
</s>
<s>
Ri(t)=Round_FO trip_NN1 time_NNT1 forith_NN1 flow_NN1 at_II timet_NN1 ._. 
</s>
<s>
Such_DA a_AT1 variant_NN1 may_VM make_VVI the_AT method_NN1 more_RGR competitive_JJ with_IW alternative_JJ algorithms_NN2 when_CS judged_VVN by_II the_AT number_NN1 of_IO batches_NN2 required_VVN ._. 
</s>
<s>
It_PPH1 is_VBZ best_JJT to_TO find_VVI the_AT "_" knee_NN1 "_" in_II the_AT load_NN1 versus_II supernode_NN1 count_NN1 curve_NN1 ,_, and_CC use_VV0 that_CST in_II determining_VVG the_AT number_NN1 of_IO supernodes_NN2 ._. 
</s>
<s>
Our_APPGE suggestion_NN1 is_VBZ to_TO exploit_VVI the_AT benefits_NN2 of_IO HTTP_NP1 as_II a_AT1 uniform_JJ communication_NN1 interface_NN1 for_IF component-based_NN1 and_CC service-based_JJ modelling_NN1 frameworks_NN2 ._. 
</s>
<s>
In_BCL21 order_BCL22 to_TO change_VVI a_AT1 convention_NN1 ,_, agents_NN2 need_VV0 to_TO re-coordinate_VVI to_II another_DD1 alternative_NN1 ._. 
</s>
<s>
A_AT1 model_NN1 of_IO ACD_NP1 has_VHZ been_VBN developed_VVN based_VVN on_II formal_JJ notation_NN1 used_VVN for_IF building_NN1 safety_NN1 critical_JJ systems_NN2 delegation_NN1 ._. 
</s>
<s>
On_II the_AT basis_NN1 of_IO these_DD2 tests_NN2 ,_, informant_NN1 bias_NN1 does_VDZ not_XX appear_VVI to_TO be_VBI a_AT1 concern_NN1 in_II this_DD1 study_NN1 ._. 
</s>
<s>
We_PPIS2 back_VV0 u_ZZ1 our_APPGE analysis_NN1 with_IW em_PPHO2 irical_JJ results_NN2 using_VVG Bayesian_JJ networks_NN2 (_( BNs_NP2 )_) ,_, both_RR synthetic_JJ and_CC from_II a_AT1 lications_NN2 ._. 
</s>
<s>
Moreover_RR ,_, NN_NP1 models_NN2 from_II the_AT formulation_NN1 based_VVN on_II Eq_NN1 ._. 
</s>
<s>
(_( 11_MC )_) shows_VVZ better_JJR representation_NN1 of_IO the_AT reference_NN1 model_NN1 than_CSN the_AT one_PN1 based_VVN on_II Eq_NN1 ._. 
</s>
<s>
(_( 12_MC )_) ._. 
</s>
<s>
A_AT1 rather_RG large_JJ sample_NN1 set_NN1 of_IO data_NN was_VBDZ obtained_VVN from_II the_AT electronic_JJ data-loggers_NN2 and_CC the_AT historic_JJ records_NN2 that_CST were_VBDR collected_VVN from_II the_AT weather_NN1 station_NN1 ._. 
</s>
<s>
This_DD1 fact_NN1 alone_RR motivates_VVZ the_AT need_NN1 for_IF adversarial-aware_JJ classifiers_NN2 ,_, that_REX21 is_REX22 ,_, algorithms_NN2 factoring_VVG in_II the_AT possibility_NN1 of_IO an_AT1 intelligent_JJ adversary_NN1 manipulating_VVG the_AT input_NN1 ._. 
</s>
<s>
Specifically_RR ,_, in_II Section4.1_FO ,_, we_PPIS2 devise_VV0 a_AT1 deterministic_JJ online_JJ algorithm_NN1 that_CST is_VBZ at_RR21 most_RR22 a_AT1 factor_NN1 of_IO 2_MC in_II cost_NN1 from_II the_AT optimal_JJ offline_JJ solution_NN1 ._. 
</s>
<s>
Adaptation_NN1 support_NN1 for_IF existing_JJ programs_NN2 has_VHZ been_VBN of_IO much_DA1 interest_NN1 at_II different_JJ levels_NN2 of_IO the_AT software_NN1 stack_NN1 in_II many_DA2 computer_NN1 science_NN1 domains_NN2 ._. 
</s>
<s>
Dependency_NN1 links_NN2 (_( )_) model_NN1 relationships_NN2 between_II actors_NN2 ._. 
</s>
<s>
Finally_RR ,_, it_PPH1 meant_VVD that_CST there_EX was_VBDZ a_AT1 need_NN1 to_TO provide_VVI a_AT1 unified_JJ namespace_NN1 page34_FO that_DD1 made_VVD cross-references_NN2 across_II the_AT various_JJ definitions_NN2 residing_VVG in_II each_DD1 of_IO the_AT specifications_NN2 ._. 
</s>
<s>
The_AT lack_NN1 of_IO explanatory_JJ power_NN1 also_RR supports_VVZ the_AT assertion_NN1 that_CST intention_NN1 should_VM not_XX be_VBI used_VVN as_II a_AT1 surrogate_NN1 for_IF actual_JJ behavior_NN1 unless_CS its_APPGE explanatory_JJ power_NN1 has_VHZ been_VBN vetted_VVN within_II the_AT associated_JJ adoption_NN1 context_NN1 (_( Chandon_NP1 et_RA21 al._RA22 ,_, 2005_MC )_) ._. 
</s>
<s>
It_PPH1 is_VBZ stated_VVN formally_RR as_CSA ,_, given_VVN wk+1=_FO (_( i_ZZ1 ,_, j_ZZ1 )_) then_RT ,_, where_RRQ and_CC ._. 
</s>
<s>
Here_RL we_PPIS2 use_VV0 D_ZZ1 (_( u_ZZ1 ,_, v_ZZ1 )_) and_CC R_ZZ1 (_( u_ZZ1 ,_, v_ZZ1 )_) to_TO denote_VVI the_AT disparity_NN1 value_NN1 and_CC the_AT reliability_NN1 label_NN1 obtained_VVN for_IF pixel_NN1 (_( u_ZZ1 ,_, v_ZZ1 )_) of_IO a_AT1 given_JJ image_NN1 ,_, with_IW R_ZZ1 (_( u_ZZ1 ,_, v_ZZ1 )_) =1_FO indicating_VVG the_AT disparity_NN1 value_NN1 D_ZZ1 (_( u_ZZ1 ,_, v_ZZ1 )_) being_VBG reliable_JJ and_CC R_ZZ1 (_( u_ZZ1 ,_, v_ZZ1 )_) =0_FO otherwise_RR ._. 
</s>
<s>
However_RR ,_, as_CSA argued_VVN by_II several_DA2 researchers_NN2 &lsqb;_( 14_MC &rsqb;_) ,_, deep_JJ architecture_NN1 ,_, composed_VVD of_IO multiple_JJ levels_NN2 of_IO non-linear_JJ operations_NN2 &lsqb;_( 15_MC &rsqb;_) ,_, is_VBZ expected_VVN to_TO perform_VVI well_RR in_II semi-supervised_JJ learning_NN1 ._. 
</s>
<s>
A_AT1 policy_NN1 can_VM also_RR be_VBI classified_VVN according_II21 to_II22 S_ZZ1 and_CC R._NP1 We_PPIS2 eliminate_VV0 merging_VVG directions_NN2 whose_DDQGE bounding_JJ boxes_NN2 overlap_VV0 with_IW the_AT other_JJ neighboring_JJ cells_NN2 ._. 
</s>
<s>
In_II contrast_NN1 ,_, if_CS the_AT combination_NN1 of_IO subgroup_NN1 priority_NN1 vectors_NN2 is_VBZ equally_RR weighted_VVN then_RT the_AT simple_JJ average_JJ coherence_NN1 value_NN1 over_II the_AT set_NN1 will_VM be_VBI a_AT1 more_RGR accurate_JJ and_CC appropriate_JJ measure_NN1 to_TO use_VVI ._. 
</s>
<s>
Here_RL ,_, we_PPIS2 see_VV0 that_CST the_AT minimum_JJ classifier_NN1 will_VM tend_VVI to_TO underestimate_VVI the_AT ground_NN1 truth_NN1 ,_, while_CS the_AT maximum_JJ classifier_NN1 will_VM tend_VVI to_TO overestimate_VVI ._. 
</s>
<s>
However_RR ,_, as_CSA the_AT case_NN1 of_IO a_AT1 conductor_NN1 with_IW a_AT1 handle_NN1 is_VBZ of_IO greater_JJR practical_JJ relevance_NN1 this_DD1 shall_VM be_VBI our_APPGE focus_NN1 ._. 
</s>
<s>
Section_NN1 2_MC introduces_VVZ the_AT MMMFEM_NN1 method_NN1 and_CC its_APPGE step-by-step_JJ formulation_NN1 leading_VVG to_II its_APPGE original_JJ implementation_NN1 ._. 
</s>
<s>
For_REX21 example_REX22 ,_, musical_JJ aspects_NN2 such_II21 as_II22 chroma_NN1 and_CC melody_NN1 may_VM be_VBI captured_VVN ,_, or_CC structural_JJ elements_NN2 like_II beats_NN2 and_CC sections_NN2 may_VM be_VBI described_VVN ._. 
</s>
<s>
Rather_II21 than_II22 trying_VVG only_RR a_AT1 single_JJ HTTP_NP1 tunnel_NN1 ,_, our_APPGE approach_NN1 was_VBDZ to_TO train_VVI a_AT1 classifier_NN1 on_II a_AT1 diverse_JJ range_NN1 of_IO HTTP_NP1 tunnels_NN2 to_TO give_VVI it_PPH1 the_AT broadest_JJT possible_JJ detection_NN1 capability_NN1 ._. 
</s>
<s>
The_AT record_NN1 includes_VVZ the_AT key_JJ phrase_NN1 that_CST triggered_VVD the_AT ad_NN1 ,_, number_NN1 of_IO impressions_NN2 for_IF that_DD1 phrase_NN1 on_II that_DD1 day_NNT1 ,_, the_AT number_NN1 of_IO clicks_NN2 ,_, the_AT average_JJ CPC_NP1 ,_, the_AT number_NN1 of_IO conversions_NN2 (_( or_CC orders_NN2 )_) ,_, the_AT total_JJ sales_NN revenues_NN2 ,_, and_CC the_AT total_JJ number_NN1 of_IO items_NN2 ordered_VVD ._. 
</s>
<s>
In_II such_DA organizations_NN2 ,_, most_RGT software-related_JJ positions_NN2 demand_VV0 multiple_JJ capabilities_NN2 ,_, including_II intra-personal_JJ ,_, organizational_JJ ,_, inter-personal_JJ and_CC management_NN1 skills_NN2 ._. 
</s>
<s>
In_II the_AT experiments_NN2 ,_, we_PPIS2 investigated_VVD a_AT1 number_NN1 of_IO different_JJ values_NN2 of_IO k_ZZ1 ._. 
</s>
<s>
Many_DA2 U.S._NP1 firms_NN2 are_VBR forced_VVN by_II privacy_NN1 laws_NN2 to_TO report_VVI privacy_NN1 breach_NN1 incidents_NN2 ,_, resulting_VVG in_II negative_JJ publicity_NN1 and_CC heightened_JJ public_JJ awareness_NN1 ._. 
</s>
<s>
Several_DA2 computational_JJ models_NN2 have_VH0 been_VBN proposed_VVN to_TO understand_VVI the_AT role_NN1 of_IO social_JJ influence_NN1 in_II obesity_NN1 &lsqb;_( 4_MC &rsqb;_) and_CC &lsqb;_( 35_MC &rsqb;_) ._. 
</s>
<s>
Savic_NP1 (_( 2002_MC )_) demonstrates_VVZ some_DD shortcomings_NN2 of_IO single-objective_JJ optimization_NN1 approaches_NN2 and_CC uses_VVZ a_AT1 multi-objective_NN1 based_VVN genetic_JJ algorithm_NN1 (_( Fonseca_NP1 and_CC Fleming_NP1 ,_, 1993_MC )_) to_TO avoid_VVI these_DD2 difficulties_NN2 ._. 
</s>
<s>
The_AT use_NN1 of_IO an_AT1 information_NN1 system_NN1 designed_VVN to_TO promote_VVI a_AT1 particular_JJ behavior_NN1 (_( such_II21 as_II22 best_JJT practice_NN1 exchange_NN1 ,_, such_II21 as_II22 inter-departmental_JJ sharing_NN1 of_IO information_NN1 ,_, such_II21 as_II22 monitoring_VVG real-time_JJ information_NN1 ,_, such_II21 as_II22 regular_JJ communication_NN1 )_) will_VM ,_, over_II time_NNT1 ,_, alter_VV0 the_AT underlying_JJ value_NN1 associated_VVN with_IW the_AT behavior_NN1 ._. 
</s>
<s>
That_REX21 is_REX22 ,_, usage_NN1 pattern_NN1 reported_VVN on_II a_AT1 Wednesday_NPD1 at_II 17:00_MC would_VM only_RR be_VBI compared_VVN to_II usage_NN1 pattern_NN1 history_NN1 on_II previous_JJ Wednesdays_NPD2 at_II around_II the_AT same_DA time_NNT1 ._. 
</s>
<s>
One_MC1 approach_NN1 to_II stepping_VVG an_AT1 audience_NN1 through_II a_AT1 model_NN1 is_VBZ to_TO present_VVI a_AT1 sequence_NN1 of_IO slides_NN2 in_II which_DDQ the_AT model_NN1 '_NULL s_ZZ1 construction_NN1 is_VBZ depicted_VVN with_IW simple_JJ shapes_NN2 ,_, and_CC each_DD1 step_NN1 is_VBZ paired_VVN with_IW details_NN2 and_CC real_JJ examples_NN2 that_CST ground_VVD the_AT conceptual_JJ form_NN1 of_IO the_AT system_NN1 in_II ways_NN2 familiar_JJ to_II the_AT audience_NN1 (_( see_VV0 Box_NN1 3_MC )_) ._. 
</s>
<s>
In_II 2003_MC ,_, a_AT1 major_JJ golf-specialty_JJ chain-store_NN1 opened_VVD a_AT1 24,000@sq_FO ._. 
</s>
<s>
However_RR ,_, in_II households_NN2 where_RRQ aggregate_JJ data_NN is_VBZ more_RGR limited_JJ or_CC where_CS overlapping_JJ appliance_NN1 usage_NN1 is_VBZ more_RGR common_JJ ,_, we_PPIS2 would_VM expect_VVI that_CST a_AT1 smaller_JJR step_NN1 size_NN1 would_VM allow_VVI a_AT1 greater_JJR number_NN1 of_IO signatures_NN2 to_TO be_VBI extracted_VVN ._. 
</s>
<s>
However_RR ,_, the_AT distinction_NN1 between_II large_JJ and_CC small_JJ violations_NN2 is_VBZ not_XX always_RR clearfor_NN1 example_NN1 ,_, sharing_VVG a_AT1 password_NN1 may_VM seem_VVI like_II a_AT1 small_JJ violation_NN1 ;_; however_RR ,_, sharing_VVG a_AT1 password_NN1 with_IW a_AT1 malicious_JJ coworker_NN1 could_VM have_VHI great_JJ consequences_NN2 ._. 
</s>
<s>
Zhang_VV0 and_CC Babovic_NN1 (_( 2012_MC )_) use_VV0 a_AT1 ROs_NP1 approach_NN1 to_TO evaluate_VVI different_JJ water_NN1 technologies_NN2 in_II water_NN1 supply_NN1 systems_NN2 under_II uncertainty_NN1 ._. 
</s>
<s>
This_DD1 shows_VVZ that_CST linearly_RR implicit_JJ Runge–Kutta_NN1 methods_NN2 with_IW step-size_JJ control_NN1 are_VBR a_AT1 good_JJ choice_NN1 for_IF solving_NN1 (_( 3_MC )_) ._. 
</s>
<s>
Examples_NN2 of_IO the_AT secondary_JJ control_NN1 masters_NN2 ,_, presented_VVN in_II the_AT Section6DSM_FO example_NN1 ,_, are_VBR the_AT consumer_NN1 and_CC the_AT electricity_NN1 retailer_NN1 ._. 
</s>
<s>
Naturally_RR ,_, some_DD types_NN2 of_IO intrusions_NN2 known_VVN for_IF their_APPGE ability_NN1 to_TO disguise_VVI the_AT behavior_NN1 as_CSA normal_JJ remain_VV0 hidden_VVN from_II our_APPGE system_NN1 unless_CS their_APPGE intrusive_JJ behavior_NN1 will_VM cause_VVI program_NN1 to_TO behave_VVI differently_RR ,_, e.g._REX ,_, mimicry_JJ attacks_NN2 ,_, buffer_NN1 overflow_NN1 attacks_NN2 ._. 
</s>
<s>
Finally_RR ,_, we_PPIS2 present_VV0 our_APPGE conclusions_NN2 in_II Section6_FO ._. 
</s>
<s>
Therefore_RR ,_, this_DD1 step_NN1 cancels_VVZ potentially_RR discriminant_JJ image_NN1 information_NN1 ._. 
</s>
<s>
For_REX21 example_REX22 ,_, &lsqb;_( 125_MC &rsqb;_) presented_VVD a_AT1 2-layer_JJ model_NN1 (_( based_VVN on_II control_NN1 flow_NN1 graphs_NN2 )_) for_IF the_AT white-box_JJ testing_NN1 of_IO web_NN1 applications_NN2 ,_, in_II which_DDQ the_AT model_NN1 had_VHD to_TO be_VBI manually_RR provided_VVN by_II the_AT user_NN1 (_( tester_NN1 )_) ._. 
</s>
<s>
Rather_RR ,_, they_PPHS2 transmit_VV0 a_AT1 pattern_NN1 of_IO neural_JJ impulses_NN2 that_CST describe_VV0 the_AT location_NN1 of_IO the_AT stimulus_NN1 and_CC such_DA properties_NN2 as_CSA its_APPGE luminance_NN1 ,_, color_NN1 ,_, shape_NN1 ,_, and_CC motion_NN1 ._. 
</s>
<s>
While_CS it_PPH1 was_VBDZ not_XX possible_JJ to_TO use_VVI experimental_JJ data_NN due_JJ to_TO lack_VVI of_IO availability_NN1 of_IO commercial_JJ MEMS_NN2 chemical_JJ sensor_NN1 arrays_NN2 ,_, experimentally_RR acquired_VVN diffusion_NN1 parameters_NN2 were_VBDR used_VVN in_II31 conjunction_II32 with_II33 the_AT currently_RR accepted_VVN mathematical_JJ model_NN1 in_II the_AT literature_NN1 ._. 
</s>
<s>
This_DD1 figure_NN1 represents_VVZ the_AT deployment_NN1 latencies_NN2 over_II the_AT course_NN1 of_IO 500_MC iterations_NN2 for_IF the_AT total_JJ deployment_NN1 latency_NN1 and_CC the_AT two_MC most_RRT time_VV0 consuming_VVG phases_NN2 :_: plan_VV0 preparation_NN1 and_CC start_VV0 launch_NN1 ._. 
</s>
<s>
The_AT SP.R3_FO Goal_NN1 is_VBZ contained_VVN in_II the_AT Field_NN1 Agent_NN1 Actor_NN1 ._. 
</s>
<s>
The_AT spatially_RR explicit_JJ FATELAND_NN1 model_NN1 (_( Pausas_NP2 ,_, 2006_MC )_) represents_VVZ species_NN competition_NN1 in_II grid_NN1 cells_NN2 as_II a_AT1 function_NN1 of_IO their_APPGE life-history_JJ characteristics_NN2 and_CC the_AT fire_NN1 regime_NN1 ._. 
</s>
<s>
Although_CS the_AT measured_JJ correlations_NN2 are_VBR not_XX that_DD1 strong_JJ ,_, previous_JJ work_NN1 showed_VVD that_CST low_JJ correlations_NN2 do_VD0 not_XX necessarily_RR mean_VVI low_JJ prediction_NN1 accuracy_NN1 ._. 
</s>
<s>
Similarly_RR ,_, much_DA1 of_IO the_AT recent_JJ development_NN1 in_II information_NN1 security_NN1 comes_VVZ from_II legal_JJ and_CC regulatory_JJ frameworks_NN2 (_( Sundt_NP1 ,_, 2006_MC )_) ,_, which_DDQ might_VM bring_VVI the_AT practice_NN1 closer_RRR to_II the_AT knowledge_NN1 domains_NN2 of_IO law_NN1 and_CC audit_NN1 ._. 
</s>
<s>
However_RR ,_, IPython_NP1 (_( Perez_NP1 and_CC Granger_NP1 ,_, 2007_MC )_) was_VBDZ found_VVN to_TO scale_VVI to_II highly_RR parallel_JJ processing_NN1 on_II heterogeneous_JJ clusters_NN2 ._. 
</s>
<s>
The_AT Transparent_JJ Authentication_NN1 Framework_NN1 uses_VVZ device_NN1 confidence_NN1 ,_, the_AT probability_NN1 that_CST the_AT current_JJ user_NN1 is_VBZ indeed_RR its_APPGE owner_NN1 ,_, to_TO authorize_VVI the_AT execution_NN1 of_IO particular_JJ tasks_NN2 ._. 
</s>
<s>
Both_DB2 NLtoSTD-BB_NP1 V2.0_FO and_CC the_AT fault_NN1 checklist_NN1 were_VBDR equally_RR effective_JJ for_IF detecting_VVG MF_NP1 faults_NN2 ._. 
</s>
<s>
This_DD1 is_VBZ particularly_RR important_JJ when_CS dealing_VVG with_IW end_NN1 users_NN2 such_II21 as_II22 subscribers_NN2 ._. 
</s>
<s>
To_TO obtain_VVI a_AT1 suitable_JJ number_NN1 of_IO frequencies_NN2 ,_, the_AT R2_FO coefficient_NN1 associated_VVN with_IW a_AT1 harmonic_JJ regression_NN1 model_NN1 is_VBZ computed_VVN ._. 
</s>
<s>
Therefore_RR ,_, the_AT proposed_JJ approach_NN1 applies_VVZ to_II both_DB2 categories_NN2 :_: requirement_NN1 modifications_NN2 and_CC database_NN1 refactoring_VVG ._. 
</s>
<s>
Our_APPGE experience_NN1 indicates_VVZ that_CST a_AT1 sensible_JJ range_NN1 of_IO is_VBZ &lsqb;_( 0.2_MC ,_, 0.4_MC &rsqb;_) ,_, for_IF it_PPH1 gives_VVZ more_DAR weights_NN2 to_II minDist_NN1 and_CC tends_VVZ to_TO create_VVI cohesive_JJ clusters_NN2 ._. 
</s>
<s>
The_AT proof_NN1 of_IO linear_JJ and_CC angular_JJ momentum_NN1 balance_NN1 for_IF the_AT three-field_JJ formulation_NN1 is_VBZ valid_JJ for_IF the_AT two-field_JJ model_NN1 with_IW p1p1_FO ,_, because_CS vv_NNU is_VBZ retained_VVN as_II an_AT1 independent_JJ field_NN1 ._. 
</s>
<s>
The_AT strength_NN1 of_IO this_DD1 model_NN1 is_VBZ its_APPGE ability_NN1 to_TO accurately_RR capture_VVI high_JJ order_NN1 nonlinearities_NN2 (_( including_II harmonics_NN2 )_) generated_VVD during_II the_AT laser–molecule_NN1 interaction_NN1 ._. 
</s>
<s>
Key_JJ weaknesses_NN2 that_CST may_VM arise_VVI from_II using_VVG interviews_NN2 are_VBR :_: bias_NN1 due_II21 to_II22 poorly_JJ constructed_JJ questions_NN2 ,_, response_NN1 bias_NN1 ,_, inaccuracies_NN2 due_II21 to_II22 poor_JJ recall_NN1 and_CC reflexivity_NN1 where_CS the_AT respondent_NN1 says_VVZ what_DDQ the_AT interviewer_NN1 expects_VVZ or_CC wants_VVZ to_TO hear_VVI &lsqb;_( 132_MC &rsqb;_) ._. 
</s>
<s>
Executives_NN2 and_CC boundary_NN1 spanners_NN2 (_( e.g._REX ,_, sales_NN agents_NN2 and_CC purchasing_VVG managers_NN2 )_) develop_VV0 outside-in_RL and_CC spanning_VVG capabilities_NN2 by_II assuming_VVG cross-organizational_JJ roles_NN2 (_( &lsqb;_( Montealegre_NP1 ,_, 2002_MC &rsqb;_) ,_, &lsqb;_( Sambamurthy_NP1 et_RA21 al._RA22 ,_, 2003_MC &rsqb;_) and_CC &lsqb;_( Levina_NP1 and_CC Vaast_NP1 ,_, 2005_MC &rsqb;_) )_) through_II which_DDQ they_PPHS2 develop_VV0 embedded_JJ relationships_NN2 with_IW outside_JJ organizations_NN2 ._. 
</s>
<s>
For_IF a_AT1 distance_NN1 metric_JJ d_ZZ1 ,_, k(t)≥d(t)/2_FU ._. 
</s>
<s>
Within_II the_AT range_NN1 of_IO profiles_NN2 examined_VVD ,_, the_AT total_JJ violations_NN2 incurred_VVN by_II the_AT planning_NN1 approach_NN1 are_VBR always_RR lower_JJR than_CSN those_DD2 incurred_VVN by_II the_AT HTC_NP1 approach_NN1 ._. 
</s>
<s>
From_II Table_NN1 14_MC ,_, we_PPIS2 see_VV0 that_CST several_DA2 hybrid_JJ test_NN1 suites_NN2 perform_VV0 better_RRR than_CSN the_AT corresponding_JJ Pure_JJ Reduction_NN1 criterion_NN1 as_CSA indicated_VVN by_II the_AT large_JJ number_NN1 of_IO bold-faced_JJ ModAPFD.C_NP1 values_NN2 ._. 
</s>
<s>
This_DD1 step_NN1 is_VBZ repeated_VVN until_CS all_DB the_AT sensors_NN2 are_VBR assigned_VVN to_II the_AT iSets_NN2 ._. 
</s>
<s>
The_AT view_NN1 that_CST command_VV0 and_CC control_NN1 will_VM improve_VVI because_CS the_AT environment_NN1 is_VBZ entirely_RR made_VVN up_RP of_IO IT_NN1 must_VM be_VBI challenged_VVN ,_, however_RR ._. 
</s>
<s>
As_CSA mentioned_VVN in_II Section3_FO ,_, our_APPGE algorithm_NN1 uses_VVZ the_AT equivalent_JJ notation_NN1 for_IF each_DD1 multicast_NN1 session_NNT1 as_II a_AT1 set_NN1 of_IO arcs_NN2 ,_, which_DDQ we_PPIS2 henceforth_RT refer_VV0 to_II as_II the_AT set_NN1 representation_NN1 of_IO the_AT session_NNT1 ._. 
</s>
<s>
Once_RR this_DD1 weight_NN1 becomes_VVZ zero_MC it_PPH1 never_RR becomes_VVZ one_MC1 in_II the_AT rest_NN1 of_IO the_AT simulation_NN1 ._. 
</s>
<s>
This_DD1 file_NN1 consists_VVZ of_IO (_( a_ZZ1 )_) the_AT access_NN1 control_NN1 policy_NN1 specification_NN1 (_( both_RR file_VV0 and_CC memory_NN1 access_NN1 control_NN1 )_) and_CC (_( b_ZZ1 )_) the_AT filenames_NN2 of_IO process_NN1 belonging_VVG to_II the_AT Process_NN1 Resident_NN1 Set_NN1 (_( PRS_NP2 )_) ._. 
</s>
<s>
This_DD1 enabled_VVD the_AT achievement_NN1 of_IO a_AT1 justifiable_JJ and_CC manageable_JJ theoretical_JJ sampling_NN1 strategy_NN1 &lsqb;_( 22_MC &rsqb;_) to_TO select_VVI the_AT data_NN and_CC define_VVI it_PPH1 within_II relevant_JJ episodes_NN2 ,_, which_DDQ would_VM support_VVI this_DD1 investigation_NN1 of_IO TUM_NN1 in_II GVTs_NP1 ._. 
</s>
<s>
The_AT shot_NN1 at_II the_AT top_NN1 of_IO the_AT stem_NN1 of_IO the_AT fork_NN1 is_VBZ the_AT video_NN1 that_CST the_AT user_NN1 is_VBZ currently_RR viewing_VVG ,_, with_IW the_AT tines_NN2 representing_VVG the_AT different_JJ threads_NN2 ._. 
</s>
<s>
The_AT reason_NN1 for_IF this_DD1 is_VBZ at_RR21 least_RR22 partially_RR due_II21 to_II22 the_AT introduction_NN1 of_IO the_AT PTPF_NP1 ._. 
</s>
<s>
The_AT index_NN1 y_ZZ1 will_VM range_VVI from_II &lsqb;_( 1_MC1 ,_, xdegree1_FO &rsqb;_) where_RRQ xdegree_NN1 is_VBZ the_AT fanout_NN1 or_CC degree_NN1 of_IO node_NN1 x_ZZ1 ._. 
</s>
<s>
We_PPIS2 assume_VV0 that_CST each_DD1 representative_JJ user_NN1 will_VM activate_VVI at_II some_DD point_NN1 of_IO time_NNT1 all_DB the_AT roles_NN2 that_CST he_PPHS1 is_VBZ authorized_VVN for_IF ._. 
</s>
<s>
Another_DD1 security_NN1 concern_NN1 of_IO biometrics_NN1 is_VBZ that_CST once_RR biometric_JJ data_NN are_VBR compromised_VVN ,_, the_AT effect_NN1 will_VM be_VBI forever_RT ._. 
</s>
<s>
We_PPIS2 are_VBR reliant_JJ on_II the_AT assessments_NN2 reported_VVN in_II &lsqb;_( 2_MC &rsqb;_) ,_, which_DDQ were_VBDR performed_VVN over_II a_AT1 period_NN1 of_IO time_NNT1 (_( several_DA2 mmonths_NN2 )_) and_CC this_DD1 could_VM be_VBI a_AT1 threat_NN1 to_II validity_NN1 because_CS the_AT assessors_NN2 may_VM have_VHI changed_VVN their_APPGE criteria_NN2 in_II this_DD1 time_NNT1 ._. 
</s>
<s>
Based_VVN on_II the_AT seminal_JJ WALKSAT_NN1 architecture_NN1 &lsqb;_( 44,45_MC &rsqb;_) ,_, we_PPIS2 introduce_VV0 a_AT1 sim_NN1 le_FW but_CCB general_JJ SLS_NP1 algorithm_NN1 called_VVN SIM_NP1 LESLS_NP2 ._. 
</s>
<s>
In_II 2007_MC part_NN1 of_IO Volume_NN1 16_MC ,_, Issue_VV0 4_MC was_VBDZ devoted_JJ to_II mobile_JJ technologies_NN2 and_CC that_DD1 part_NN1 of_IO the_AT issue_NN1 was_VBDZ edited_VVN by_II M._NN1 Rossi_NN1 ,_, V.K._NP1 Tuunainen_NP1 and_CC S._NP1 Jarvenpaa_NP1 ._. 
</s>
<s>
Other_JJ types_NN2 of_IO censoring_VVG can_VM occur_VVI as_RR21 well_RR22 (_( see_VV0 Collett_NP1 ,_, 2003_MC ,_, pp_NN1 ._. 
</s>
<s>
Trajectory_NN1 Repulsion_NN1 begins_VVZ by_II evaluating_VVG the_AT objective_JJ function_NN1 on_II a_AT1 single_JJ uniform_JJ random_JJ sample_NN1 of_IO points_NN2 and_CC discarding_VVG those_DD2 points_NN2 for_IF which_DDQ the_AT objective_JJ function_NN1 is_VBZ above_II the_AT median_NN1 ._. 
</s>
<s>
The_AT scenario_NN1 consists_VVZ of_IO 6_MC SPARQL_NN1 queries_NN2 denoted_VVD Q1–6_NP1 ._. 
</s>
<s>
Another_DD1 reason_NN1 for_IF choosing_VVG OpenMP_NP1 was_VBDZ that_CST it_PPH1 is_VBZ supported_VVN by_II most_DAT compilers_NN2 and_CC has_VHZ matured_VVN into_II a_AT1 well-understood_JJ and_CC standardized_JJ parallel_JJ computing_NN1 API_NN2 for_IF which_DDQ a_AT1 lot_NN1 of_IO performance_NN1 and_CC testing_VVG tools_NN2 are_VBR available_JJ ._. 
</s>
<s>
The_AT questionnaires_NN2 were_VBDR self-report_JJ ,_, Likert-type_JJ scale_NN1 instruments_NN2 measuring_VVG the_AT participants_NN2 '_NULL attitudes_NN2 toward_II their_APPGE simulation_NN1 game_NN1 play_NN1 experience_NN1 ._. 
</s>
<s>
We_PPIS2 have_VH0 also_RR run_VVN several_DA2 simulations_NN2 of_IO 1000_MC LB_NNU steps_NN2 where_CS we_PPIS2 render_VV0 and_CC write_VV0 an_AT1 image_NN1 to_II disk_NN1 every_AT1 5–200_MCMC LB_NNU steps_NN2 ._. 
</s>
<s>
In_II the_AT second_MD iteration_NN1 ,_, nodep2is_FO added_VVN to_TO set_VVI of_IO vertices_NN2 of_IO graphG_NN1 and_CC the_AT edge_NN1 (_( p1_FO ,_, p2_FO )_) is_VBZ added_VVN to_TO set_VVI of_IO edges_NN2 ofG_VV0 ._. 
</s>
<s>
Further_RRR ,_, calling_VVG onG_NN1 will_VM result_VVI inTGhaving_VVG the_AT same_DA topology_NN1 asG_NN1 ,_, since_CS the_AT number_NN1 of_IO nodes_NN2 ofG_VV0 is_VBZ two_MC ._. 
</s>
<s>
The_AT differences_NN2 between_II the_AT median_JJ simulated_JJ data_NN and_CC SOM_NNU1 predictions_NN2 indicates_VVZ areas_NN2 with_IW higher_JJR uncertainties_NN2 or_CC anomalous_JJ values_NN2 to_TO be_VBI further_RRR investigated_VVN regarding_II issues_NN2 of_IO overexploitation_NN1 by_II pumping_NN1 ,_, potential_JJ hydraulic_JJ connections_NN2 with_IW the_AT underlying_JJ Guarani_NNU sandstone_NN1 aquifer_NN1 ,_, and_CC surface_NN1 contamination_NN1 by_II agricultural_JJ compounds_NN2 ._. 
</s>
<s>
Here_RL the_AT consultancy_NN1 becomes_VVZ a_AT1 part_NN1 of_IO the_AT innovative_JJ solution_NN1 itself_PPX1 ._. 
</s>
<s>
Fig._NN1 3_MC shows_VVZ the_AT emissions_NN2 volatility_NN1 (_( )_) and_CC the_AT mean_JJ and_CC maximum_JJ monthly_JJ emissions_NN2 (_( inset_NN1 )_) as_II a_AT1 function_NN1 of_IO m_ZZ1 ._. 
</s>
<s>
For_IF each_DD1 value_NN1 of_IO m_ZZ1 ,_, 10_MC market_NN1 simulations_NN2 were_VBDR carried_VVN out_RP using_VVG different_JJ random_JJ initial_JJ conditions_NN2 ._. 
</s>
<s>
Further_RRR ,_, our_APPGE focus_NN1 in_II this_DD1 work_NN1 was_VBDZ on_II core_NN1 developers_NN2 ,_, as_II21 against_II22 the_AT team_NN1 ._. 
</s>
<s>
Liles_NP1 et_RA21 al_RA22 ._. 
</s>
<s>
argue_VV0 that_CST the_AT objective_JJ principle_NN1 can_VM be_VBI applied_VVN to_II cyber_NN1 warfare_NN1 without_IW much_DA1 work_NN1 ;_; those_DD2 engaged_VVN in_II cyber_NN1 warfare_NN1 will_VM have_VHI objectives_NN2 ,_, and_CC launch_VV0 attacks_NN2 to_TO achieve_VVI those_DD2 objectives_NN2 ._. 
</s>
<s>
Management_NN1 and_CC control_NN1 functions_NN2 for_IF nodes_NN2 in_II InstaGENI_JJ racks_NN2 are_VBR primarily_RR provided_VVN by_II the_AT ProtoGENI_JJ software_NN1 stack_NN1 ._. 
</s>
<s>
A_AT1 Bayesian_JJ network_NN1 is_VBZ used_VVN to_TO model_VVI the_AT detecting_NN1 sensors_NN2 and_CC their_APPGE interdependence_NN1 ._. 
</s>
<s>
With_IW Sordo_NP1 method_NN1 ,_, when_CS K_ZZ1 is_VBZ equal_JJ to_II the_AT size_NN1 of_IO the_AT training_NN1 set_NN1 ,_, precision_NN1 is_VBZ reduced_VVN to_II what_DDQ is_VBZ achieved_VVN by_II assigning_VVG the_AT most_RGT popular_JJ tags_NN2 in_II the_AT collection_NN1 ._. 
</s>
<s>
Further_JJR research_NN1 is_VBZ needed_VVN to_TO position_VVI switching_JJ cost_NN1 within_II larger_JJR nomonological_JJ networks_NN2 of_IO behavior_NN1 ._. 
</s>
<s>
Ostrand_VV0 et_RA21 al_RA22 ._. 
</s>
<s>
were_VBDR able_JJ with_IW high_JJ accuracy_NN1 to_TO predict_VVI the_AT number_NN1 of_IO faults_NN2 in_II files_NN2 in_II two_MC large_JJ industrial_JJ systems_NN2 &lsqb;_( 51_MC &rsqb;_) ._. 
</s>
<s>
Since_CS D_ZZ1 is_VBZ fixed_VVN ,_, we_PPIS2 will_VM henceforth_RT drop_VVI the_AT subscript_NN1 and_CC simply_RR write_VV0 R_ZZ1 for_IF the_AT regression_NN1 operator_NN1 ._. 
</s>
<s>
The_AT proof_NN1 for_IF this_DD1 proposition_NN1 is_VBZ given_VVN in_II ?_? 
</s>
<s>
Appendix_NN1 A._NNU SfM-based_JJ methods_NN2 can_VM be_VBI categorized_VVN into_II dense_JJ correspondence-based_JJ methods_NN2 and_CC sparse_JJ correspondence-based_JJ methods_NN2 according_II21 to_II22 the_AT density_NN1 of_IO the_AT corresponding_JJ 2D_NNU FFPs_NP1 ._. 
</s>
<s>
Such_DA explorations_NN2 could_VM provide_VVI insights_NN2 into_II the_AT peculiarities_NN2 of_IO software_NN1 team_NN1 dynamics_NN ,_, inform_VV0 appropriate_JJ team_NN1 configurations_NN2 ,_, and_CC enable_VV0 the_AT early_JJ identification_NN1 of_IO '_NULL software_NN1 gems_NN2 '_NULL exceptional_JJ practitioners_NN2 in_II31 terms_II32 of_II33 both_RR task_NN1 and_CC team_NN1 performance_NN1 ._. 
</s>
<s>
If_CS the_AT super-net_NN1 has_VHZ not_XX been_VBN announced_VVN withinhprefixtime_NNT1 ,_, then_RT the_AT sub-prefix_JJ hijack_NN1 will_VM fail_VVI to_TO be_VBI detected_VVN ._. 
</s>
<s>
According_II21 to_II22 the_AT Straussian_JJ approach_NN1 of_IO data_NN analysis_NN1 for_IF grounded_JJ theory_NN1 ,_, prior_JJ knowledge_NN1 acquired_VVN through_II the_AT literature_NN1 and/or_CC previous_JJ studies_NN2 can_VM inform_VVI future_JJ research_NN1 productively_RR ._. 
</s>
<s>
It_PPH1 is_VBZ the_AT fastest_JJT method_NN1 on_II coarser_JJR domain_NN1 decompositions_NN2 ,_, but_CCB Method_NN1 2_MC outperforms_VVZ it_PPH1 slightly_RR on_II 27_MC or_CC more_DAR subdomains_NN2 ._. 
</s>
<s>
Perdisci_NP1 et_RA21 al_RA22 ._. 
</s>
<s>
(_( 2009_MC )_) developed_VVD McPAD_NP1 to_TO detect_VVI shellcode_NN1 attacks_NN2 on_II web_NN1 servers_NN2 ._. 
</s>
<s>
A_AT1 fourth_MD runs_VVZ FlowVisor_NN1 to_TO provide_VVI support_NN1 for_IF control-plane_NN1 multi-tenancy_NN1 on_II the_AT OpenFlow_NN1 network_NN1 ._. 
</s>
<s>
When_CS the_AT negation_NN1 is_VBZ applied_VVN to_II a_AT1 literal_JJ ,_, double_JJ negations_NN2 are_VBR implicitly_RR removed_VVN ,_, that_REX21 is_REX22 ,_, iflis_NN1 pthen_NN1 lisp_NN1 ._. 
</s>
<s>
One_MC1 of_IO the_AT fundamental_JJ problems_NN2 of_IO SfM-based_JJ methods_NN2 is_VBZ self-occlusion_JJ which_DDQ means_VVZ some_DD facial_JJ parts_NN2 occlude_VV0 other_JJ facial_JJ parts_NN2 when_RRQ head_NN1 rotation_NN1 occurs_VVZ ._. 
</s>
<s>
Note_VV0 that_DD1 Fig._NN1 2_MC shows_NN2 state_VV0 transitions_NN2 of_IO the_AT AC_NN1 with_II31 respect_II32 to_II33 a_AT1 single_JJ auction_NN1 ._. 
</s>
<s>
The_AT parent_NN1 vessel_NN1 had_VHD a_AT1 nominal_JJ radius_NN1 of_IO 0.67cm_NNU and_CC a_AT1 length_NN1 of_IO 13.86cm_NNU ._. 
</s>
<s>
Over_II a_AT1 collection_NN1 interval_NN1 of_IO 15_MC min_NNU ,_, statistical_JJ measures_NN2 are_VBR calculated_VVN to_TO form_VVI MCD_MC features_NN2 such_II21 as_II22 :_: the_AT number_NN1 of_IO new_JJ source-destination_JJ pairs_NN2 seen_VVN ,_, and_CC the_AT number_NN1 of_IO new_JJ source-destination_JJ pairs_NN2 which_DDQ are_VBR not_XX in_II the_AT long_JJ term_NN1 database_NN1 ._. 
</s>
<s>
For_REX21 example_REX22 ,_, the_AT Annotea_NP1 approach_NN1 to_II locating_VVG an_AT1 annotation_NN1 at_II a_AT1 particular_JJ point_NN1 in_II a_AT1 document_NN1 uses_VVZ XPointers_NN2 ._. 
</s>
<s>
We_PPIS2 denote_VV0 the_AT resulting_JJ proof_NN1 of_IO knowledge_NN1 of_IO the_AT plaintextvi(t)_NN1 byPoKvi(t)_NN1 ._. 
</s>
<s>
Legitimate_JJ user_NN1 behavior_NN1 :_: the_AT operations_NN2 on_II file_NN1 and_CC application_NN1 content_NN1 ,_, the_AT level_NN1 of_IO legitimate_JJ user_NN1 sophistication_NN1 in_II31 terms_II32 of_II33 his/her_APPGE IT_NN1 knowledge_NN1 and_CC the_AT network_NN1 I/O_NN1 operations_NN2 of_IO the_AT user_NN1 are_VBR evaluated_VVN using_VVG suitable_JJ metrics_NN2 ._. 
</s>
<s>
A_AT1 combination_NN1 of_IO personality_NN1 traits_NN2 helps_VVZ to_II narrow_JJ the_AT discrepancy_NN1 between_II intention_NN1 and_CC behavior_NN1 by_II increasing_VVG predictive_JJ ability_NN1 of_IO intention_NN1 on_II user_NN1 '_NULL s_ZZ1 behavior_NN1 (_( Conner_NP1 and_CC Abraham_NP1 ,_, 2001_MC ,_, Courneya_NP1 et_RA21 al._RA22 ,_, 1999_MC and_CC Rhodes_NP1 and_CC Courneya_NP1 ,_, 2003_MC )_) ._. 
</s>
<s>
Increased_JJ perception_NN1 of_IO usefulness_NN1 are_VBR likely_JJ to_TO result_VVI from_II increased_JJ ability_NN1 to_TO sort_VVI data_NN ,_, generate_VV0 reports_NN2 ,_, compile_VV0 graphs_NN2 ,_, attach_VV0 documents_NN2 to_TO email_VVI ,_, etc_RA ._. 
</s>
<s>
Encryption_NN1 techniques_NN2 could_VM also_RR hide_VVI the_AT transport_NN1 of_IO cyber_NN1 weapons_NN2 between_II seller_NN1 and_CC buyer_NN1 ._. 
</s>
<s>
The_AT future_NN1 of_IO these_DD2 frameworks_NN2 very_RG much_DA1 depends_VVZ upon_II how_RGQ wide_JJ their_APPGE standards_NN2 will_VM be_VBI accepted_VVN within_II the_AT modeling_NN1 community_NN1 and_CC whether_CSW there_EX will_VM appear_VVI a_AT1 critical_JJ mass_NN1 of_IO contributed_VVN models_NN2 ._. 
</s>
<s>
This_DD1 approach_NN1 is_VBZ indeed_RR the_AT only_JJ approach_NN1 possible_JJ when_CS noa_NN1 prioriinformation_NN1 is_VBZ available_JJ about_II the_AT behavior_NN1 of_IO the_AT system_NN1 ._. 
</s>
<s>
The_AT replication-aware_JJ distributed_JJ evaluation_NN1 of_IO a_AT1 SPARQL_NN1 query_NN1 Q_ZZ1 over_II an_AT1 arbitrary_JJ graph_NN1 cover_NN1 called_VVN cover_VV0 that_CST assigns_VVZ triples_NN2 of_IO an_AT1 arbitrary_JJ RDF_NP1 graph_NN1 G_ZZ1 to_TO compute_VVI nodes_NN2 C_ZZ1 ,_, denoted_VVN by_II Q_ZZ1 cover_VV0 ,_, is_VBZ defined_VVN as_CSA Q_ZZ1 cover_VV0 |_NULL ,_, C∈_FO c∈C_FO Q_ZZ1 coverc_NN1 ._. 
</s>
<s>
These_DD2 projects_NN2 share_VV0 a_AT1 common_JJ theme_NN1 with_IW our_APPGE work_NN1 in_II that_DD1 centralized_JJ design_NN1 of_IO adaptation_NN1 strategies_NN2 can_VM be_VBI specified_VVN at_II a_AT1 high_JJ level_NN1 for_IF existing_JJ programs_NN2 ._. 
</s>
<s>
Competition_NN1 for_IF water_NN1 and_CC light_JJ following_JJ disturbance_NN1 (_( such_II21 as_II22 wildfire_NN1 )_) and_CC along_II gradients_NN2 of_IO these_DD2 resources_NN2 is_VBZ the_AT predominant_JJ cause_NN1 of_IO characteristic_JJ Mediterranean_JJ community_NN1 structures_NN2 (_( &lsqb;_( Vila_NP1 and_CC Sardans_NP2 ,_, 1999_MC &rsqb;_) and_CC &lsqb;_( Zavala_NP1 et@al._FO ,_, 2000_MC &rsqb;_) )_) ._. 
</s>
<s>
In_II focus_NN1 group_NN1 2_MC ,_, there_EX were_VBDR 2_MC design_NN1 vulnerability_NN1 pathways_NN2 ,_, 3_MC implementation_NN1 vulnerability_NN1 pathways_NN2 ,_, 2_MC configuration_NN1 vulnerability_NN1 pathways_NN2 ,_, and_CC 2_MC operational_JJ vulnerability_NN1 pathways_NN2 ._. 
</s>
<s>
An_AT1 example_NN1 of_IO this_DD1 approach_NN1 is_VBZ the_AT CUAHSI_JJ HydroDesktop_NN1 application_NN1 that_CST provides_VVZ access_NN1 to_II remote_JJ data_NN archives_NN2 made_VVD available_JJ using_VVG the_AT CUAHSI_JJ WaterOneFlow_NN1 web_NN1 service_NN1 (_( Ames_NP1 et_NP1 ?_? al._NNU ,_, 2012_MC ;_; Tarboton_NP1 et_NP1 ?_? al._NNU ,_, 2009_MC )_) ._. 
</s>
<s>
Out_II21 of_II22 these_DD2 criteria_NN2 ,_, the_AT Canny_JJ edge_NN1 detector_NN1 was_VBDZ developed_VVN ,_, which_DDQ is_VBZ probably_RR the_AT most_RGT widely_RR used_JJ detector_NN1 and_CC considered_VVD to_TO be_VBI the_AT standard_JJ edge_NN1 detection_NN1 algorithm_NN1 in_II the_AT computer_NN1 vision_NN1 applications_NN2 &lsqb;_( 19_MC &rsqb;_) ._. 
</s>
<s>
These_DD2 two_MC implementation_NN1 decisions_NN2 are_VBR made_VVN to_TO allow_VVI quick_JJ development_NN1 of_IO the_AT prototype_NN1 of_IO our_APPGE framework_NN1 ._. 
</s>
<s>
Even_CS21 though_CS22 the_AT final_JJ simulation_NN1 time_NNT1 is_VBZ T=8000_FO only_RR the_AT data_NN from_II k=2001_FO to_II k=8000_FO is_VBZ accounted_VVN for_IF in_BCL21 order_BCL22 to_TO remove_VVI the_AT initial_JJ transient_JJ effects_NN2 caused_VVN by_II the_AT setups_NN2 of_IO the_AT PSN_NP1 model_NN1 always_RR with_IW empty_JJ queues_NN2 ._. 
</s>
<s>
In_BCL21 order_BCL22 to_TO identify_VVI foundational_JJ principles_NN2 of_IO cyber_NN1 security_NN1 ,_, a_AT1 strategic_JJ research_NN1 approach_NN1 is_VBZ absolutely_RR essential_JJ ._. 
</s>
<s>
However_RR ,_, these_DD2 differences_NN2 were_VBDR not_XX elaborated_VVN further_RRR ,_, and_CC there_EX is_VBZ no_AT reference_NN1 to_II a_AT1 situation_NN1 where_RRQ a_AT1 strategic_JJ IT_NN1 application_NN1 is_VBZ outsourced_VVN ._. 
</s>
<s>
Observation_NN1 is_VBZ a_AT1 time_NNT1 consuming_VVG process_NN1 ._. 
</s>
<s>
In_RR21 short_RR22 ,_, veiled_JJ certificates_NN2 improve_VV0 personal_JJ privacy_NN1 by_II limiting_JJ database_NN1 cross-linking_NN1 ,_, thereby_RR reducing_VVG the_AT risk_NN1 of_IO identity_NN1 theft_NN1 ._. 
</s>
<s>
These_DD2 costs_NN2 may_VM include_VVI direct_JJ expenses_NN2 or_CC relate_VV0 to_II human_JJ resource_NN1 investments_NN2 and/or_CC acquisitions_NN2 of_IO durable_JJ assets_NN2 (_( i.e._REX ,_, machines_NN2 ,_, production_NN1 facilities_NN2 )_) (_( Spekman_NP1 and_CC Strauss_NP1 ,_, 1986_MC )_) ._. 
</s>
<s>
Applications_NN2 that_CST integrate_VV0 data_NN of_IO some_DD source_NN1 systems_NN2 are_VBR at_II risk_NN1 of_IO inconsistency_NN1 occurring_VVG in_II the_AT aggregated_JJ results_NN2 ._. 
</s>
<s>
Although_CS anti-forensic_JJ procedures_NN2 are_VBR usually_RR designed_VVN to_TO destroy_VVI evidence_NN1 (_( Distefano_NP1 et_RA21 al._RA22 ,_, 2010_MC )_) ,_, those_DD2 which_DDQ are_VBR intended_VVN not_XX to_TO leave_VVI evidence_NN1 ,_, enhancing_VVG the_AT user_NN1 '_NULL s_ZZ1 confidentiality_NN1 ,_, must_VM also_RR be_VBI included_VVN in_II the_AT category_NN1 ,_, and_CC are_VBR the_AT focus_NN1 of_IO this_DD1 article_NN1 ._. 
</s>
<s>
Features_NN2 that_CST made_VVD this_DD1 scenario_NN1 Navy-specific_NN1 included_VVD the_AT protection_NN1 of_IO classified_JJ information_NN1 and_CC cultural_JJ aspects_NN2 of_IO organizational_JJ security_NN1 associated_VVN with_IW the_AT hierarchical_JJ command_NN1 structure_NN1 of_IO the_AT DoD_NP1 ._. 
</s>
<s>
The_AT CP_NP1 and_CC CP+M_FO approaches_NN2 have_VH0 100%_NNU recall_VV0 by_II to_II their_APPGE extensive_JJ approximations_NN2 ._. 
</s>
<s>
Thef–N_VV0 ?_? relationship_NN1 reported_VVD herein_RR will_VM not_XX be_VBI representative_JJ of_IO the_AT repeated_JJ citation_NN1 characteristics_NN2 of_IO a_AT1 forward-chronological_JJ search_NN1 ._. 
</s>
<s>
MnM_NP1 was_VBDZ designed_VVN to_II mark-up_NN1 training_NN1 data_NN for_IF IE_REX tools_NN2 rather_II21 than_II22 as_II an_AT1 annotation_NN1 tool_NN1 per_RR21 se_RR22 &lsqb;_( 36_MC &rsqb;_) ._. 
</s>
<s>
The_AT social_JJ and_CC political_JJ skills_NN2 required_VVN to_TO interact_VVI with_IW demanding_VVG executives_NN2 were_VBDR seen_VVN as_CSA quite_RG distinct_JJ from_II the_AT technical_JJ process_NN1 of_IO risk_NN1 mitigation_NN1 ._. 
</s>
<s>
The_AT IG_NN1 heuristic_JJ solutions_NN2 were_VBDR always_RR inferior_JJ ,_, and_CC these_DD2 results_NN2 are_VBR hardly_RR surprising_JJ ._. 
</s>
<s>
The_AT graph_NN1 edit_VV0 distance_NN1 therefore_RR provides_VVZ a_AT1 well-defined_JJ way_NN1 of_IO measuring_VVG the_AT similarity_NN1 of_IO two_MC graphs_NN2 ._. 
</s>
<s>
This_DD1 is_VBZ required_VVN for_IF design_NN1 purposes_NN2 as_II an_AT1 initial_JJ set_NN1 of_IO experimental_JJ snapshots_NN2 may_VM not_XX be_VBI sufficient_JJ ._. 
</s>
<s>
However_RR ,_, fault_NN1 tolerance_NN1 and_CC search_NN1 latency_NN1 tradeoff_NN1 with_IW efficiency_NN1 ,_, since_CS redundancy_NN1 results_NN2 in_II extra_JJ work_NN1 for_IF peers_NN2 ._. 
</s>
<s>
The_AT following_JJ theorem_NN1 shows_VVZ that_CST the_AT reduction_NN1 is_VBZ indeed_RR correct_JJ ._. 
</s>
<s>
CX-DIFF_NP1 detects_VVZ customized_JJ changes_NN2 on_II XML_NP1 documents_NN2 like_II keywords_NN2 and_CC phrases_NN2 ._. 
</s>
<s>
The_AT goal_NN1 is_VBZ to_TO predict_VVI the_AT fuel_NN1 consumption_NN1 from_II the_AT characteristics_NN2 of_IO the_AT car_NN1 ,_, i.e._REX ,_, a_AT1 regression_NN1 model_NN1 has_VHZ to_TO be_VBI learned_VVN ._. 
</s>
<s>
These_DD2 are_VBR the_AT methods_NN2 that_CST are_VBR exposed_VVN to_II both_RR network_NN1 and_CC application-layer_JJ components_NN2 ._. 
</s>
<s>
REE_NP1 is_VBZ a_AT1 function_NN1 of_IO the_AT percentage_NN1 of_IO lean_JJ and_CC fat_JJ mass_NN1 which_DDQ we_PPIS2 approximated_VVD as_II a_AT1 fixed_JJ percentage_NN1 of_IO body_NN1 weight_NN1 ._. 
</s>
<s>
See_VV0 for_REX21 example_REX22 &lsqb;_( 4_MC &rsqb;_) ,_, &lsqb;_( 29_MC &rsqb;_) ,_, &lsqb;_( 30_MC &rsqb;_) ,_, &lsqb;_( 17_MC &rsqb;_) and_CC &lsqb;_( 31_MC &rsqb;_) ._. 
</s>
<s>
It_PPH1 is_VBZ not_XX very_RG helpful_JJ in_II finding_NN1 (_( hierarchical_JJ )_) routes_NN2 in_II a_AT1 constantly_RR changing_JJ network_NN1 like_II the_AT vehicle_NN1 grid_NN1 unless_CS it_PPH1 is_VBZ combined_VVN with_IW the_AT Mobile_NP1 IP_NP1 construct_VV0 ,_, with_IW provides_VVZ the_AT desired_JJ redirection_NN1 ._. 
</s>
<s>
The_AT single_JJ vulnerability_NN1 found_VVN in_II the_AT 14_MC yields_VVZ a_AT1 vulnerabilities_NN2 discovered_VVD per_II hour_NNT1 value_NN1 of_IO 0.07_MC ._. 
</s>
<s>
If_CS the_AT framework_NN1 operates_VVZ in_II client-server_JJ mode_NN1 ,_, the_AT communication_NN1 engine_NN1 works_VVZ as_II a_AT1 bridge_NN1 between_II the_AT capture_NN1 device_NN1 and_CC the_AT comprehensive_JJ framework_NN1 ._. 
</s>
<s>
The_AT timed_JJ behavior_NN1 set_NN1 of_IO TH_NP1 is_VBZ defined_VVN as_II a_AT1 set_NN1 which_DDQ contains_VVZ all_DB possible_JJ run_NN1 traces_NN2 of_IO complete_JJ paths_NN2 taken_VVN by_II TH_NP1 in_II ,_, i.e._REX ._. 
</s>
<s>
can_VM be_VBI thought_VVN of_IO as_II a_AT1 set_NN1 of_IO all_DB possible_JJ timed_JJ behaviors_NN2 of_IO TH_NP1 which_DDQ can_VM affect_VVI the_AT response_NN1 time_NNT1 of_IO TL_NP1 ._. 
</s>
<s>
Antoniol_VV0 et_RA21 al._RA22 &lsqb;_( 2_MC &rsqb;_) classified_JJ issues_NN2 reported_VVN into_II bugs_NN2 and_CC non-bugs_NN2 ,_, where_CS a_AT1 "_" bug_NN1 "_" referred_VVN to_II a_AT1 corrective_JJ maintenance_NN1 request_NN1 and_CC other_JJ requests_NN2 (_( "_" non-bugs_NN2 "_" )_) referred_VVD to_II perfective_JJ and_CC adaptive_JJ maintenance_NN1 ,_, refactoring_VVG ,_, discussions_NN2 ,_, requests_NN2 for_IF help_NN1 ,_, etc_RA ._. 
</s>
<s>
Upon_II receipt_NN1 of_IO this_DD1 message_NN1 ,_, x2_FO sets_VVZ threshold_NN1 =_FO 4_MC (_( line_NN1 8_MC )_) ._. 
</s>
<s>
We_PPIS2 have_VH0 also_RR investigated_VVN the_AT effects_NN2 of_IO user-fatigue_NN1 ,_, how_RRQ this_DD1 can_VM be_VBI ameliorated_VVN by_II appropriate_JJ algorithm_NN1 design_NN1 ,_, and_CC demonstrated_VVD other_JJ novel_JJ methods_NN2 for_IF countering_VVG this_DD1 &lsqb;_( 4_MC &rsqb;_) ._. 
</s>
<s>
However_RR ,_, the_AT target_NN1 vegetation_NN1 that_CST may_VM develop_VVI under_II the_AT '_NULL Room_NN1 for_IF the_AT River_NN1 '_NULL plans_NN2 will_VM eventually_RR lead_VVI to_II a_AT1 higher_JJR –_- but_CCB yet_RR harder_JJR to_TO predict_VVI –_- hydraulic_JJ roughness_NN1 ,_, and_CC thus_RR an_AT1 increase_NN1 in_II water_NN1 levels_NN2 (_( Makaske_NP1 et_NP1 ?_? al._NNU ,_, 2011_MC )_) ._. 
</s>
<s>
If_CS this_DD1 common_JJ cyclic_JJ pattern_NN1 is_VBZ removed_VVN from_II an_AT1 individual_JJ '_NULL s_ZZ1 datastream_VV0 ,_, the_AT remaining_JJ information_NN1 describes_VVZ the_AT individual_NN1 '_NULL s_ZZ1 uniqueness_NN1 or_CC difference_NN1 to_II the_AT population_NN1 norm_NN1 ._. 
</s>
<s>
This_DD1 section_NN1 provides_VVZ some_DD necessary_JJ technical_JJ details_NN2 and_CC some_DD background_NN1 to_TO motivate_VVI the_AT computational_JJ problems_NN2 investigated_VVN in_II the_AT paper_NN1 ._. 
</s>
<s>
A_AT1 survey_NN1 was_VBDZ conducted_VVN of_IO purchasing/procurement_NN1 professionals_NN2 in_II the_AT manufacturing_NN1 sector_NN1 to_TO test_VVI the_AT hypotheses_NN2 presented_VVN above_RL ._. 
</s>
<s>
Return_VV0 the_AT tilted-time_JJ windows_NN2 in_II which_DDQ patternpoccurred_VVD with_IW support1_FO ,_, where1>=_FO (_( 1is_FO the_AT specified_JJ support_NN1 in_II the_AT query_NN1 andis_VVZ the_AT support_NN1 of_IO the_AT incremental_JJ mining_NN1 framework_NN1 )_) ._. 
</s>
<s>
We_PPIS2 also_RR provide_VV0 recommendations_NN2 for_IF software_NN1 project_NN1 governance_NN1 and_CC show_VV0 how_RRQ the_AT outcomes_NN2 of_IO our_APPGE work_NN1 have_VH0 implications_NN2 for_IF team_NN1 strategies_NN2 ._. 
</s>
<s>
There_EX are_VBR three_MC such_DA code_NN1 lists_NN2 ._. 
</s>
<s>
Whilst_CS AHP_NP1 serves_VVZ the_AT function_NN1 of_IO making_VVG priorities_NN2 between_II criteria_NN2 explicit_JJ ,_, DST_NP1 enables_VVZ a_AT1 unified_JJ decision_NN1 to_TO be_VBI made_VVN by_II fusing_VVG the_AT opinions_NN2 of_IO multiple_JJ stakeholders_NN2 to_II a_AT1 single_JJ measure_NN1 of_IO performance_NN1 for_IF each_DD1 criterion_NN1 ._. 
</s>
<s>
As_II21 of_II22 its_APPGE nature_NN1 ,_, the_AT diffusion_NN1 process_NN1 will_VM eventually_RR stop_VVI and_CC will_VM reach_VVI equilibrium_NN1 state_NN1 ,_, where_CS either_RR the_AT complete_JJ networks_NN2 is_VBZ adopted_VVN or_CC no_AT more_RGR new_JJ adopters_NN2 are_VBR reachable_JJ through_II existing_JJ links_NN2 ._. 
</s>
<s>
InTable_JJ 2_MC ,_, note_VV0 how_RRQ most_DAT of_IO the_AT high_JJ expertise_NN1 behaviors_NN2 appear_VV0 in_II the_AT intentional_JJ destruction_NN1 category_NN1 ._. 
</s>
<s>
Here_RL the_AT existing_JJ centralized_JJ water_NN1 supply_NN1 system_NN1 is_VBZ incapable_JJ of_IO meeting_VVG the_AT increasing_JJ user_NN1 demands_NN2 for_IF the_AT entire_JJ planning_NN1 horizon_NN1 ,_, even_CS21 if_CS22 it_PPH1 is_VBZ allowed_VVN that_CST no_AT water_NN1 is_VBZ supplied_VVN to_II any_DD user_NN1 nodes_NN2 during_II the_AT failure_NN1 period_NN1 ._. 
</s>
<s>
Hence_RR ,_, the_AT verifier_NN1 will_VM perform_VVI either_RR 1_MC1 orksignature_NN1 operations_NN2 to_TO validate_VVI the_AT prefixes_NN2 ._. 
</s>
<s>
They_PPHS2 may_VM aim_VVI for_IF competitive_JJ advantage_NN1 or_CC simply_RR parity_NN1 within_II their_APPGE industry_NN1 ,_, where_CS with_IW rapid_JJ change_NN1 and_CC turbulence_NN1 ,_, many_DA2 inevitably_RR play_VV0 catch-up_NN1 (_( see_VV0 ,_, e.g._REX ,_, McAfee_NP1 and_CC Brynjolfsson_NP1 ,_, 2008_MC )_) ._. 
</s>
<s>
The_AT three-body_JJ potential_NN1 diverges_VVZ when_RRQ →_NULL -0_MC ,_, with_IW bonds_NN2 otherwise_RR unstretched_JJ ,_, causing_VVG the_AT bond_NN1 angles_NN2 to_TO remain_VVI in_II the_AT interval_NN1 (_( -+0_MC ,_, -0_MC )_) ._. 
</s>
<s>
Finally_RR ,_, an_AT1 (_( again_RT ,_, manual_NN1 ,_, text-only-based_NN1 )_) assessment_NN1 of_IO the_AT '_NULL success_NN1 '_NULL of_IO the_AT application_NN1 was_VBDZ made_VVN ._. 
</s>
<s>
Some_DD participants_NN2 who_PNQS could_VM not_XX borrow_VVI enough_DD resources_NN2 would_VM not_XX be_VBI able_JK to_TO achieve_VVI their_APPGE desired_JJ QoS_NN2 level_NN1 ._. 
</s>
<s>
The_AT present_JJ paper_NN1 focuses_VVZ on_II long-duration_JJ failures_NN2 that_CST would_VM deplete_VVI local_JJ storage_NN1 before_II repair_NN1 is_VBZ completed_VVN ._. 
</s>
<s>
Finally_RR ,_, in_II order_NN1 to_TO further_RRR illustrate_VVI the_AT difference_NN1 between_II mapping_VVG studies_NN2 and_CC SLRs_NP1 ,_, in_II Table_NN1 2_MC ,_, we_PPIS2 compare_VV0 Jrgensen_NP1 and_CC Shepperd_NP1 '_NULL s_ZZ1 mapping_VVG study_NN1 with_IW Kitchenham_NP1 et_RA21 al_RA22 ._. 
</s>
<s>
'_NULL s_ZZ1 SLR_NP1 &lsqb;_( 30_MC &rsqb;_) ,_, using_VVG on_II the_AT criteria_NN2 adopted_VVN in_II Table_NN1 1_MC1 plus_II an_AT1 additional_JJ criterion_NN1 "_" Recommendations_NN2 "_" ._. 
</s>
<s>
Step2-3_MC Hyper-Mutation_NN1 :_: All_DB the_AT antibodies_NN2 in_II the_AT temporary_JJ population_NN1 undergo_VV0 the_AT hyper-mutation_NN1 ._. 
</s>
<s>
It_PPH1 employs_VVZ a_AT1 simplified_JJ tree-shaped_JJ query_NN1 representation_NN1 and_CC distributes_VVZ functionality_NN1 to_II different_JJ widgets_NN2 with_II31 respect_II32 to_II33 the_AT accord_NN1 between_II the_AT functionality_NN1 ,_, interaction_NN1 ,_, and_CC representation_NN1 paradigms_NN2 (_( i.e._REX ,_, tabular_JJ widget_NN1 for_IF template_NN1 selection_NN1 ,_, and_CC menu-based_JJ widget_NN1 for_IF navigation_NN1 )_) ._. 
</s>
<s>
Previous_JJ versions_NN2 of_IO IPAT_NP1 used_VVD a_AT1 "_" standard_NN1 "_" (_( 1+5_FO )_) evolution_NN1 strategy_NN1 with_IW isotropic_JJ mutation_NN1 ,_, so_CS changes_NN2 were_VBDR equally_RR likely_JJ to_TO happen_VVI in_II either_DD1 direction_NN1 ._. 
</s>
<s>
Under_II an_AT1 imputation_NN1 in_II the_AT -core_NN1 ,_, the_AT excess_NN1 e(C)=v(C)_FO p(C)_NNU of_IO any_DD coalition_NN1 C_ZZ1 is_VBZ at_RR21 most_RR22 ._. 
</s>
<s>
Sandpiper_NN1 '_NULL s_ZZ1 CPU_NN1 and_CC network_NN1 overhead_RL is_VBZ dependent_JJ on_II the_AT number_NN1 of_IO PM_NP1 s_ZZ1 and_CC VM_NP1 s_ZZ1 in_II the_AT data_NN center_NN1 ._. 
</s>
<s>
According_II21 to_II22 Table_NN1 10_MC ,_, MK-SVD_NP1 is_VBZ the_AT fastest_JJT one_PN1 among_II the_AT four_MC compared_VVD methods_NN2 ,_, which_DDQ also_RR gains_VVZ relative_JJ satisfactory_JJ denoising_JJ results_NN2 for_IF most_DAT cases_NN2 in_II our_APPGE experiments_NN2 ._. 
</s>
<s>
Similar_JJ to_II GRAIL_NN1 ,_, Secure_VV0 Tropos_NP1 provides_VVZ a_AT1 single-user_JJ perspective_NN1 on_II goal_NN1 refinement_NN1 ,_, whereas_CS our_APPGE compliance_NN1 framework_NN1 incorporates_VVZ multiple_JJ viewpoints_NN2 through_II distributed_JJ refinement_NN1 ._. 
</s>
<s>
The_AT fundamental_JJ idea_NN1 for_IF this_DD1 NLtoSTD-BB_JJ transformation_NN1 is_VBZ that_CST a_AT1 functional_JJ requirement_NN1 should_VM typically_RR describe_VVI an_AT1 entity_NN1 transitioning_VVG from_II one_MC1 state_NN1 to_II another_DD1 ._. 
</s>
<s>
You_PPY decide_VV0 to_TO fix_VVI this_DD1 by_II changing_VVG the_AT Domain_NN1 to_II PizzaIngredient_NP1 ._. 
</s>
<s>
More_RGR technical_JJ descriptions_NN2 of_IO the_AT robot_NN1 may_VM be_VBI found_VVN in_II previous_JJ papers_NN2 ._. 
</s>
<s>
IEM_VV0 concepts_NN2 and_CC early_JJ models_NN2 are_VBR now_RT more_DAR than_CSN thirty_MC years_NNT2 old_JJ (_( Bailey_NP1 et_NP1 ?_? al._NNU ,_, 1985_MC ;_; Cohen_NP1 ,_, 1986_MC ;_; Mackay_NP1 ,_, 1991_MC ;_; Meadows_NN2 et_NP1 ?_? al._NNU ,_, 1972_MC ;_; Walters_NP ,_, 1986_MC )_) ._. 
</s>
<s>
Unfortunately_RR ,_, although_CS digital_JJ forensic_JJ readiness_NN1 (_( DFR_NP1 )_) is_VBZ becoming_VVG a_AT1 legal_JJ and_CC regulatory_JJ requirement_NN1 in_II many_DA2 jurisdictions_NN2 in_II the_AT western_JJ world_NN1 ,_, studies_NN2 show_VV0 that_CST most_DAT organisations_NN2 especially_RR in_II Australia_NP1 have_VH0 not_XX developed_VVN a_AT1 significant_JJ capability_NN1 in_II this_DD1 domain_NN1 (_( e.g._REX the_AT Australian_JJ Institute_NN1 of_IO Criminology_NP1 reports_VVZ that_CST less_DAR than_CSN 2%_NNU of_IO Australian_JJ organizations_NN2 have_VH0 a_AT1 plan_NN1 for_IF digital_JJ forensics_NN2 ,_, see_VV0 AIC_NP1 (_( 2009_MC )_) )_) ._. 
</s>
<s>
Conceptually_RR ,_, all_DB mesodata_NN1 is_VBZ stored_VVN in_II amesodata_NN1 layer_NN1 (_( although_CS in_II practice_NN1 ,_, they_PPHS2 may_VM be_VBI split_VVN between_II other_JJ system_NN1 files_NN2 and_CC user_NN1 relations_NN2 as_CSA appropriate_JJ )_) ._. 
</s>
<s>
The_AT transition_NN1 local_JJ time_NNT1 is_VBZ labeled_VVN ?_? t_ZZ1 ?_? 
</s>
<s>
OFF_RP ._. 
</s>
<s>
We_PPIS2 revisit_VV0 the_AT capabilities_NN2 of_IO content_JJ negotiation_NN1 in_II the_AT next_MD section_NN1 ._. 
</s>
<s>
The_AT question_NN1 is_VBZ therefore_RR :_: what_DDQ kind_NN1 of_IO information_NN1 is_VBZ essential_JJ and_CC how_RRQ can_VM it_PPH1 best_RRT be_VBI passed_VVN on_II ?_? 
</s>
<s>
Conditions_NN2 (_( ii_MC )_) ,_, (_( iii_MC )_) ,_, and_CC (_( iv_MC )_) in_II (_( 10_MC )_) combines_VVZ parallel_JJ paths_NN2 while_CS discarding_VVG all_DB intersecting_VVG paths13_FO ._. 
</s>
<s>
The_AT figure_NN1 shows_VVZ that_CST the_AT bound_NN1 is_VBZ asymptotic_JJ and_CC for_REX21 example_REX22 when_CS b_ZZ1 =_FO 32_MC ,_, the_AT actual_JJ bound_NN1 reached_VVD by_RP is_VBZ about_RG 27_MC ._. 
</s>
<s>
The_AT aleatoric_JJ uncertainty_NN1 in_II a_AT1 performance_NN1 measure_NN1 is_VBZ quantified_VVN by_II its_APPGE verification_NN1 diameter(8)_FO ,_, i.e._REX ,_, the_AT largest_JJT deviation_NN1 in_II performance_NN1 that_CST is_VBZ computed_VVN when_CS each_DD1 input_NN1 parameter_NN1 is_VBZ allowed_VVN to_TO vary_VVI in_II turn_NN1 between_II pairs_NN2 of_IO values_NN2 spanning_VVG its_APPGE entire_JJ range_NN1 ._. 
</s>
<s>
Consequently_RR ,_, we_PPIS2 chose_VVD to_TO adopt_VVI correlation_NN1 analysis_NN1 as_CSA it_PPH1 could_VM be_VBI used_VVN to_TO explore_VVI the_AT complex_JJ relationships_NN2 between_II the_AT application_NN1 of_IO IS_VBZ capabilities_NN2 and_CC process-level_JJ improvements_NN2 in_II competitive_JJ positioning_NN1 ,_, at_II the_AT level_NN1 of_IO the_AT individual_JJ capability_NN1 and_CC process_NN1 ,_, as_RG well_RR as_CSA in_II their_APPGE aggregated_JJ forms_NN2 ._. 
</s>
<s>
Finally_RR ,_, in_II Section_NN1 5_MC we_PPIS2 offer_VV0 some_DD conclusions_NN2 from_II this_DD1 study_NN1 ._. 
</s>
<s>
Dynamic_JJ information_NN1 is_VBZ held_VVN in_II an_AT1 XML_NP1 database_NN1 (_( the_AT IBM_NP1 TSpaces_NP2 tuple_NN1 space_NN1 server_NN1 )_) ._. 
</s>
<s>
Training_NN1 ,_, however_RR ,_, is_VBZ dependent_JJ on_II guides_NN2 or_CC other_JJ similar_JJ materials_NN2 that_CST the_AT Field_NN1 Agent_NN1 can_VM use_VVI to_TO show_VVI the_AT HOT_JJ how_RRQ to_TO perform_VVI surveillance_NN1 and_CC that_DD1 can_VM be_VBI left_VVN for_IF use_NN1 when_CS conducting_VVG future_JJ surveys_NN2 ._. 
</s>
<s>
The_AT experimental_JJ results_NN2 reported_VVN above_RL clearly_RR demonstrate_VV0 that_CST the_AT proposed_JJ multi-manifold_NN1 based_VVN classification_NN1 approach_NN1 is_VBZ effective_JJ for_IF facial_JJ expression_NN1 recognition_NN1 ,_, and_CC it_PPH1 overwhelms_VVZ the_AT single_JJ manifold_NN1 based_VVN approaches_NN2 ._. 
</s>
<s>
This_DD1 number_NN1 gives_VVZ an_AT1 idea_NN1 of_IO the_AT level_NN1 of_IO misdirection_NN1 caused_VVN by_II overlay_NN1 routing_NN1 ,_, and_CC tends_VVZ to_TO be_VBI proportional_JJ to_II the_AT level_NN1 of_IO cross-layer_NN1 conflict&lsqb;38&rsqb;_FO ._. 
</s>
<s>
Our_APPGE approach_NN1 differs_VVZ from_II these_DD2 works_NN ,_, because_CS they_PPHS2 focus_VV0 on_II creating_VVG an_AT1 optimal_JJ reduced_JJ or_CC minimized_JJ test_NN1 suite_NN1 without_IW regard_NN1 to_II the_AT order_NN1 of_IO the_AT test_NN1 cases_NN2 in_II the_AT reduced_JJ set_NN1 ._. 
</s>
<s>
In_RR21 addition_RR22 ,_, global_JJ DRSD_NP1 has_VHZ simple_JJ calculation_NN1 ;_; it_PPH1 preserves_VVZ both_RR the_AT shape_NN1 and_CC continuity_NN1 of_IO mesh_NN1 surfaces_NN2 ;_; it_PPH1 is_VBZ one-to-one_JJ and_CC intrinsic_JJ to_II underlying_JJ surfaces_NN2 ._. 
</s>
<s>
To_TO minimise_VVI the_AT effect_NN1 of_IO the_AT match_NN1 cost_NN1 during_II the_AT γ_NULL evaluation_NN1 we_PPIS2 set_VV0 it_PPH1 to_II a_AT1 constant_JJ and_CC use_VV0 values_NN2 of_IO γ_NULL between_II 1_MC1 to_II 10_MC with_IW cross-validation_JJ and_CC NN_NP1 classification_NN1 in_BCL21 order_BCL22 to_TO find_VVI an_AT1 optimum_JJ value_NN1 ._. 
</s>
<s>
However_RR ,_, the_AT maximum_JJ mass_JJ conservation_JJ error_NN1 in_II these_DD2 cases_NN2 was3.19×10-5_FO ._. 
</s>
<s>
However_RR ,_, our_APPGE emphasis_NN1 here_RL is_VBZ not_XX on_II the_AT consequences_NN2 of_IO the_AT adversary_NN1 '_NULL s_ZZ1 actions_NN2 in_II a_AT1 real_JJ setting_NN1 ,_, but_CCB rather_RR on_II the_AT assumption_NN1 that_CST attacks_NN2 are_VBR anomalous_JJ events_NN2 which_DDQ nonetheless_RR might_VM be_VBI conveniently_RR camouflaged_VVN to_TO avoid_VVI detection_NN1 ._. 
</s>
<s>
To_TO represent_VVI each_DD1 separate_JJ element_NN1 ?_? a_AT1 ?_? of_IO the_AT set_NN1 ?_? 
</s>
<s>
A_ZZ1 ?_? in_II the_AT Bloom_NP1 filter_NN1 ,_, the_AT bits_NN2 of_IO ?_? v_ZZ1 ?_? at_II the_AT positions_NN2 ?_? h1(a)_FO ,_, ?_? h2(a)_FO ,_, ?_? ?_? ,_, ?_? hk(a)_NN1 are_VBR set_VVN to_II one_MC1 (_( thus_RR ,_, a_AT1 particular_JJ bit_NN1 of_IO a_AT1 Bloom_NN1 filter_NN1 can_VM be_VBI set_VVN to_II one_PN1 by_II several_DA2 elements_NN2 of_IO the_AT set_NN1 )_) ._. 
</s>
<s>
A_AT1 baseline_NN1 ,_, conventional_JJ multinomial_JJ LDA_NN1 is_VBZ first_MD computed_VVD ,_, giving_VVG a_AT1 test_NN1 sample_NN1 accuracy_NN1 of_IO 93.84%_FO ._. 
</s>
<s>
For_IF the_AT computation_NN1 of_IO both_DB2 kernels_NN2 ,_, the_AT RDF_NP1 graph_NN1 G_ZZ1 can_VM be_VBI restricted_VVN to_II those_DD2 vertices_NN2 that_CST are_VBR actually_RR within_II a_AT1 distance_NN1 d_ZZ1 of_IO at_RR21 least_RR22 one_MC1 instance_NN1 vertex_NN1 i_ZZ1 ._. 
</s>
<s>
Furthermore_RR ,_, in_II the_AT for_IF loop_NN1 over_II the_AT vertices_NN2 V_ZZ1 in_II Algorithms_NN2 3.1_MC and_CC 3.2_MC ,_, we_PPIS2 can_VM exclude_VVI vertices_NN2 v_ZZ1 for_IF which_DDQ i∈I:dist_NN1 (_( i_ZZ1 ,_, v_ZZ1 )_) +n≤d_FO ._. 
</s>
<s>
The_AT distance_NN1 to_TO target_VVI weights_NN2 for_IF the_AT particular_JJ case_NN1 study_NN1 used_VVN in_II this_DD1 research_NN1 are_VBR provided_VVN in_II the_AT case_NN1 study_NN1 results_NN2 within_II Section_NN1 5_MC ._. 
</s>
<s>
Fig._NN1 ?_? 1_MC1 shows_VVZ an_AT1 upper_JJ view_NN1 of_IO the_AT Bocal_JJ in_II which_DDQ it_PPH1 is_VBZ possible_JJ to_TO observe_VVI the_AT Ebro_NP1 River_NNL1 ,_, the_AT Pignatelli_JJ dam_NN1 ,_, and_CC the_AT Gate_NN1 House_NN1 ._. 
</s>
<s>
In_RR21 particular_RR22 ,_, the_AT model_NN1 does_VDZ not_XX explicitly_RR support_VVI activities_NN2 to_II ontology_NN1 maintenance_NN1 ,_, including_II the_AT evolution_NN1 of_IO the_AT instance_NN1 base_NN1 which_DDQ are_VBR incrementally_RR added_VVN to_II the_AT ontology_NN1 ._. 
</s>
<s>
Since_CS the_AT global_JJ response_NN1 is_VBZ mainly_RR governed_VVN by_II the_AT local_JJ connection_NN1 behavior_NN1 ,_, the_AT displacement_NN1 at_II the_AT second_MD floor_NN1 in_II Fig._NN1 17b_FO can_VM partly_RR check_VVI the_AT accuracy_NN1 of_IO the_AT prediction_NN1 made_VVN by_II the_AT trained_JJ NN_NP1 models_NN2 ._. 
</s>
<s>
The_AT space_NN1 mesh_NN1 controls_VVZ the_AT construction_NN1 of_IO the_AT spacetime_NNT1 mesh_NN1 ,_, comprised_VVN of_IO (_( d+1_FO )_) (_( d+1_FO )_) -simplices_NN2 ,_, as_CSA described_VVN below_RL ._. 
</s>
<s>
Second_MD ,_, the_AT workload_NN1 for_IF each_DD1 processor_NN1 may_VM become_VVI increasingly_RR unbalanced_VVN due_II21 to_II22 a_AT1 large_JJ variation_NN1 in_II the_AT number_NN1 of_IO mortar_NN1 degrees_NN2 of_IO freedom_NN1 per_II subdomain_NN1 ._. 
</s>
<s>
A_AT1 Tent_NN1 Pitcher_NN1 patch_NN1 is_VBZ a_AT1 cluster_NN1 of_IO tetrahedral_JJ elements_NN2 whose_DDQGE boundary_NN1 is_VBZ a_AT1 collection_NN1 of_IO oriented_JJ triangular_JJ facets_NN2 in_II spacetime_NNT1 ._. 
</s>
<s>
The_AT thresholds_NN2 chosen_VVN for_IF the_AT classifiers_NN2 are_VBR appropriate_JJ for_IF this_DD1 dataset_NN1 ._. 
</s>
<s>
We_PPIS2 compared_VVD the_AT observed_JJ error_NN1 rates_NN2 for_IF the_AT top-k_JJ weight_NN1 of_IO lower_VV0 bounds_NN2 obtained_VVN via_II the_AT split-sample_NN1 and_CC the_AT 2-fold_JJ methods_NN2 ._. 
</s>
<s>
Weidema_NP1 et_RA21 al_RA22 ._. 
</s>
<s>
(_( 2003_MC )_) pointed_VVD out_RP that_DD1 monitoring_NN1 of_IO emission_NN1 data_NN differs_VVZ between_II companies_NN2 ._. 
</s>
<s>
With_IW this_DD1 in_II mind_NN1 ,_, it_PPH1 is_VBZ potentially_RR confusing_JJ that_CST subsequent_JJ dialogs_NN2 still_RR use_VV0 the_AT terms_NN2 distinctly_RR ._. 
</s>
<s>
From_II these_DD2 results_NN2 we_PPIS2 can_VM draw_VVI the_AT conclusion_NN1 that_CST :_: an_AT1 appropriate_JJ spring_NN1 model_NN1 for_IF the_AT RBC_NP1 should_VM have_VHI the_AT maximum_NN1 allowed_VVN extension_NN1 length_NN1 ,_, in_II the_AT neighborhood_NN1 of_IO which_DDQ the_AT spring_NN1 force_NN1 rapidly_RR hardens_VVZ in_BCL21 order_BCL22 to_TO prevent_VVI further_JJR membrane_NN1 strain_NN1 ._. 
</s>
<s>
For_IF the_AT example_NN1 ofFig_NN1 ._. 
</s>
<s>
A.4_FO ,_, the_AT instancesI1andI2are_FO computed_VVN by_II selecting_VVG the_AT value_NN1 4_MC ,_, and_CC removing_VVG the_AT intervalx3_FO ,_, respectively_RR ._. 
</s>
<s>
We_PPIS2 also_RR maintain_VV0 that_CST other_JJ architectures_NN2 that_CST use_VV0 similar_JJ metastructures_NN2 could_VM also_RR work_VVI efficiently_RR with_IW the_AT middleware_NN1 ,_, while_CS others_NN2 might_VM require_VVI additional_JJ effort_NN1 ._. 
</s>
<s>
However_RR ,_, the_AT results_NN2 were_VBDR partly_RR inconsistent_JJ due_II21 to_II22 ambiguous_JJ interpretation_NN1 of_IO some_DD criteria_NN2 by_II different_JJ people_NN ;_; for_IF this_DD1 reason_NN1 ,_, they_PPHS2 had_VHD to_TO be_VBI eventually_RR consolidated_VVN by_II a_AT1 single_JJ person_NN1 ._. 
</s>
<s>
Boundaries_NN2 y=0y=0_FO and_CC y=1y=1_FO are_VBR Dirichlet_NN1 type_NN1 and_CC the_AT rest_NN1 of_IO the_AT boundary_NN1 is_VBZ Neumann_NP1 type_NN1 ._. 
</s>
<s>
This_DD1 is_VBZ a_AT1 piecewise_JJ technique_NN1 that_CST can_VM be_VBI reapplied_VVN at_II various_JJ locations_NN2 around_II a_AT1 simulation_NN1 ._. 
</s>
<s>
We_PPIS2 use_VV0 a_AT1 t-test_JJ to_TO assess_VVI if_CSW the_AT land-cover_JJ composition_NN1 for_IF each_DD1 set_NN1 of_IO model_NN1 repetitions_NN2 is_VBZ statistically_RR significantly_RR different_JJ from_II the_AT default_NN1 parameter_NN1 set_NN1 ._. 
</s>
<s>
We_PPIS2 calculated_VVD the_AT bias_NN1 to_TO quantify_VVI if_CSW our_APPGE model_NN1 fit_NN1 was_VBDZ on_II average_NN1 overestimating_VVG or_CC underestimating_VVG EVI_NN2 ._. 
</s>
<s>
In_BCL21 order_BCL22 to_TO assess_VVI how_RRQ the_AT precision_NN1 of_IO floating_JJ point_NN1 operations_NN2 impacts_NN2 on_II the_AT convergence_NN1 of_IO the_AT algorithm_NN1 ,_, we_PPIS2 counted_VVD the_AT number_NN1 of_IO steps_NN2 needed_VVN to_TO reach_VVI a_AT1 given_JJ accuracy_NN1 (_( imposed_VVN using_VVG the_AT convergence_NN1 criteria_NN2 )_) on_II both_RR GPU_NP1 and_CC CPU_NN1 ._. 
</s>
<s>
Finally_RR ,_, users_NN2 themselves_PPX2 may_VM have_VHI motive_NN1 to_TO compromise_VVI assets_NN2 ._. 
</s>
<s>
The_AT value_NN1 for_IF Modaily_NP1 was_VBDZ 3.07_MC ._. 
</s>
<s>
Table_NN1 5_MC @_II presents_NN2 a_AT1 list_NN1 of_IO the_AT clustering_NN1 algorithms_NN2 used_VVN in_II the_AT paper_NN1 ._. 
</s>
<s>
For_IF many_DA2 other_JJ base_NN1 semantics_NN1 ,_, we_PPIS2 show_VV0 that_DD1 complexity_NN1 of_IO ideal_JJ reasoning_NN1 coincides_VVZ with_IW the_AT complexity_NN1 of_IO skeptical_JJ acceptance_NN1 of_IO the_AT base_NN1 semantics_NN1 ._. 
</s>
<s>
Inspired_VVN by_II this_DD1 observation_NN1 ,_, geographic_JJ information_NN1 retrieval_NN1 (_( GIR_NP1 )_) systems_NN2 attempt_VV0 to_TO identify_VVI spatial_JJ constraints_NN2 in_II queries_NN2 ,_, and_CC to_TO determine_VVI which_DDQ web_NN1 resources_NN2 satisfy_VV0 them_PPHO2 &lsqb;_( 1_MC1 &rsqb;_) ,_, &lsqb;_( 2_MC &rsqb;_) ._. 
</s>
<s>
The_AT semantics_NN1 of_IO SPARQL_NN1 queries_NN2 is_VBZ given_VVN in_II31 terms_II32 of_II33 solution_NN1 mappings_NN2 ,_, which_DDQ are_VBR partial_JJ maps_NN2 s:V_FO →_NULL T_ZZ1 with_IW (_( possibly_RR empty_JJ )_) domain_NN1 dom(s)_NN2 ._. 
</s>
<s>
Then_RT the_AT master_NN1 node_NN1 selects_VVZ the_AT firstksamples_NN2 fromD_VV0 to_TO decide_VVI upon_II the_AT class_NN1 of_IO the_AT unknown_JJ sample_NN1 ._. 
</s>
<s>
In_II the_AT second_MD set_NN1 of_IO tests_NN2 we_PPIS2 fixed_VVD the_AT number_NN1 of_IO points_NN2 and_CC clusters_NN2 as_CSA before_RT ,_, but_CCB increased_VVD the_AT number_NN1 of_IO dimensions_NN2 in_II the_AT data_NN set_VVN from_II 10_MC to_II 120_MC ._. 
</s>
<s>
The_AT role_NN1 title_NN1 definitions_NN2 used_VVN for_IF systems_NN2 development_NN1 were_VBDR collated_VVN and_CC any_DD exhibiting_NN1 nine_MC or_CC more_DAR occurrences_NN2 were_VBDR retained_VVN (_( as_CSA there_EX was_VBDZ a_AT1 clear_JJ cut-off_NN1 below_II that_DD1 range_NN1 )_) ._. 
</s>
<s>
The_AT Jaccard_NP1 coefficient_NN1 measures_VVZ the_AT proportion_NN1 of_IO pairs_NN2 that_CST belong_VV0 to_II the_AT same_DA cluster_NN1 (_( a_ZZ1 )_) in_II both_DB2 partitions_NN2 ,_, relative_II21 to_II22 all_DB pairs_NN2 that_CST belong_VV0 to_II the_AT same_DA cluster_NN1 in_II at_RR21 least_RR22 one_MC1 of_IO the_AT two_MC partitions_NN2 @_II (_( a+b+c_FO )_) ._. 
</s>
<s>
It_PPH1 should_VM also_RR be_VBI noted_VVN that_CST the_AT focus_NN1 of_IO the_AT current_JJ study_NN1 was_VBDZ on_II the_AT frequency_NN1 with_IW which_DDQ references_NN2 are_VBR cited_VVN repeatedly_RR within_II journal_NN1 articles_NN2 ._. 
</s>
<s>
The_AT uncertainty_NN1 estimates_NN2 presented_VVN for_IF measured_JJ TSS_NP1 ,_, NO3-N_JJ ,_, PO4-P_JJ ,_, total_JJ N_ZZ1 ,_, and_CC total_JJ P_ZZ1 loads_NN2 and_CC concentrations_NN2 provide_VV0 fundamental_JJ information_NN1 related_VVD to_TO discharge_VVI and_CC water_NN1 quality_NN1 data_NN ._. 
</s>
<s>
Fig._NN1 3_MC provides_VVZ an_AT1 overview_NN1 of_IO this_DD1 process_NN1 ._. 
</s>
<s>
Two_MC of_IO the_AT main_JJ differences_NN2 between_II RDF_NP1 graphs_NN2 and_CC the_AT typical_JJ graphs_NN2 used_VVN in_II graph_NN1 mining_NN1 and_CC machine_VV0 learning_VVG from_II graphs_NN2 ,_, are_VBR that_DD1 vertices_VVZ in_II RDF_NP1 graphs_NN2 have_VH0 unique_JJ labels3_FO and_CC there_EX is_VBZ a_AT1 large_JJ number_NN1 of_IO different_JJ labels_NN2 overall_RR ._. 
</s>
<s>
The_AT main_JJ reason_NN1 is_VBZ the_AT following_JJ :_: both_RR AML-bk_JJ and_CC GOMMA-bk_JJ use_NN1 mapping_NN1 composition_NN1 techniques_NN2 and_CC the_AT reuse_NN1 of_IO mappings_NN2 between_II UMLS_NN2 ,_, Uberon_NP1 and_CC FMA_NP1 ._. 
</s>
<s>
Furthermore_RR ,_, at_II a_AT1 2_MC pixel_NN1 matching_NN1 boundary_NN1 ,_, if_CS the_AT genuine_JJ acceptance_NN1 rate_NN1 is_VBZ reduced_VVN to_II 99.9%_FO ,_, the_AT false_JJ acceptance_NN1 rate_NN1 can_VM be_VBI improved_VVN by_II 0.0034–0.0011%_FO ._. 
</s>
<s>
For_IF datasets_NN2 such_II21 as_II22 YAGO_NP1 ,_, 7_MC those_DD2 types_NN2 are_VBR often_RR very_RG informative_JJ ,_, for_REX21 example_REX22 ,_, products_NN2 may_VM have_VHI concise_JJ types_NN2 such_II21 as_II22 Smartphone_NP1 or_CC AndroidDevice_NP1 ._. 
</s>
<s>
The_AT usability_NN1 and_CC security_NN1 of_IO the_AT existing_JJ single-factor_NN1 ,_, knowledge-based_JJ procedure_NN1 was_VBDZ compared_VVN to_II a_AT1 two-factor_JJ approach_NN1 involving_VVG use_NN1 of_IO a_AT1 hardware_NN1 token_NN1 ._. 
</s>
<s>
The_AT 34_MC continuous_JJ features_NN2 and_CC the_AT converted_JJ binary_JJ values_NN2 were_VBDR then_RT all_RR input_VVN to_II PCA_NP1 ._. 
</s>
<s>
Heuristic_JJ :_: The_AT teacher_NN1 is_VBZ given_VVN teaching_VVG guidance_NN1 that_CST is_VBZ based_VVN on_II a_AT1 computational_JJ teaching_NN1 heuristic_JJ ._. 
</s>
<s>
When_CS Z_ZZ1 is_VBZ Cl_FO or_CC Br_JJ ,_, then_RT the_AT F-Z_NP1 VS_II ,_, max_NN1 is_VBZ more_RGR positive_JJ than_CSN the_AT NCZ_NP1 ;_; e.g._REX compare_VV0 FCl_NP1 and_CC NCCl_NP1 ._. 
</s>
<s>
Therefore_RR ,_, when_CS making_VVG methodological_JJ choices_NN2 ,_, it_PPH1 must_VM be_VBI kept_VVN in_II mind_NN1 that_CST globally_RR distributed_VVN teams_NN2 will_VM differ_VVI from_II co-located_JJ ones_NN2 ._. 
</s>
<s>
Examples_NN2 of_IO such_DA algorithms_NN2 can_VM be_VBI found_VVN ._. 
</s>
<s>
Reverse_VV0 next_MD constructs_NN2 the_AT vote_NN1 :_: p_ZZ1 c1_FO c2_FO c3_FO ._. 
</s>
<s>
The_AT dimensions_NN2 allow_VV0 to_TO describe_VVI PETs_NN2 from_II different_JJ and_CC complementary_JJ privacy-related_JJ perspectives_NN2 to_TO create_VVI a_AT1 more_RGR comprehensive_JJ view_NN1 of_IO a_AT1 PET_NN1 ._. 
</s>
<s>
Therefore_RR ,_, the_AT proposed_JJ approach_NN1 maybe_RR used_VVN in_II numerical_JJ oceanography_NN1 to_TO solve_VVI other_JJ hydrodynamic_JJ problems_NN2 ._. 
</s>
<s>
The_AT UCM_JJ notation_NN1 shares_VVZ many_DA2 characteristics_NN2 with_IW UML_NP1 activity_NN1 diagrams_NN2 but_CCB offers_VVZ more_DAR flexibility_NN1 in_II how_RRQ sub-diagrams_NN2 can_VM be_VBI connected_VVN and_CC sub-components_NN2 can_VM be_VBI represented_VVN ._. 
</s>
<s>
This_DD1 case_NN1 study_NN1 attempts_NN2 to_TO simulate_VVI the_AT last_MD 100_MC years_NNT2 of_IO events_NN2 and_CC compare_VV0 the_AT outcome_NN1 with_IW the_AT present_JJ day_NNT1 ._. 
</s>
<s>
The_AT paper_NN1 concludes_VVZ with_IW Section_NN1 8_MC ,_, which_DDQ discusses_VVZ lessons_NN2 learned_VVN and_CC Section_NN1 9_MC with_IW conclusions_NN2 and_CC directions_NN2 for_IF future_JJ work_NN1 ._. 
</s>
<s>
Next_MD ,_, we_PPIS2 briefly_RR describe_VV0 our_APPGE research_NN1 method_NN1 ,_, including_II the_AT data_NN collection_NN1 and_CC processing_VVG procedures_NN2 ;_; and_CC the_AT analytical_JJ steps_NN2 using_VVG the_AT concept_NN1 analysis_NN1 and_CC mapping_VVG software_NN1 ._. 
</s>
<s>
Moreover_RR ,_, given_JJ diagram_NN1 Band_NN1 a_AT1 specific_JJ I_ZZ1 with_IW one_MC1 object_NN1 ,_, model_NN1 evaluation_NN1 is_VBZ in_II P_ZZ1 because_CS there_EX is_VBZ only_RR one_MC1 valuation_NN1 to_TO consider_VVI ._. 
</s>
<s>
In_II other_JJ instances_NN2 ,_, for_REX21 example_REX22 the_AT relatively_RR large_JJ topic-area_NN1 1_MC1 ,_, IS_VBZ for_IF Strategic_JJ Decision_NN1 Making_VVG in_II Volume_NN1 9_MC ,_, is_VBZ in_RR21 part_RR22 explained_VVN by_II a_AT1 special_JJ edition_NN1 devoted_VVD to_II knowledge_NN1 management_NN1 issues_NN2 ._. 
</s>
<s>
Unfortunately_RR ,_, observe_VV0 that_CST ,_, no_DDQV31 matter_DDQV32 what_DDQV33 m_ZZ1 does_VDZ ,_, any_DD route_NN1 from_II pu_FO to_II m_ZZ1 that_CST has_VHZ zu_FW as_II a_AT1 next_MD hop_NN1 can_VM not_XX be_VBI of_IO length_NN1 less_DAR than_CSN 4_MC (_( in_II fact_NN1 ,_, this_DD1 is_VBZ the_AT case_NN1 even_CS21 if_CS22 m_ZZ1 hijacks_NN2 d_ZZ1 '_NULL s_ZZ1 prefix_NN1 and_CC announces_VVZ it_PPH1 to_II euv_NN1 )_) ._. 
</s>
<s>
The_AT function_NN1 codes_NN2 therefore_RR highly_RR enable_VV0 the_AT attacks_NN2 shown_VVN inFig.1_FO ,_, including_II surveillance_NN1 ,_, DoS_NN1 ,_, and_CC directly_RR controlling_VVG device_NN1 operations_NN2 ._. 
</s>
<s>
These_DD2 ,_, unlike_JJ alphanumeric_JJ passwords_NN2 and_CC codes_NN2 ,_, can_VM not_XX be_VBI written_VVN down_RP or_CC shared_VVD easily_RR ._. 
</s>
<s>
While_CS ontology-based_JJ question_NN1 answering_NN1 systems_NN2 over_II restricted_JJ domains_NN2 often_RR interpret_VV0 a_AT1 question_NN1 with_II31 respect_II32 to_II33 an_AT1 unambiguous_JJ ontology_NN1 ,_, in_II the_AT case_NN1 of_IO large_JJ open-domain_JJ ontologies_NN2 such_II21 as_II22 DBpedia_NP1 they_PPHS2 encounter_VV0 a_AT1 wide_JJ range_NN1 of_IO ambiguous_JJ wordssuddenly_RR one_MC1 query_NN1 term_NN1 can_VM have_VHI multiple_JJ interpretations_NN2 within_II the_AT same_DA ontology_NN1 ._. 
</s>
<s>
In_II Section_NN1 2_MC we_PPIS2 discuss_VV0 previous_JJ work_NN1 on_II masquerade_NN1 detection_NN1 and_CC mimicry_JJ attacks_NN2 ._. 
</s>
<s>
To_TO conclude_VVI the_AT definition_NN1 ,_, we_PPIS2 need_VV0 an_AT1 extended_JJ version_NN1 of_IO the_AT Lock_NP1 Conversion_NN1 Table_NN1 ,_, to_TO govern_VVI the_AT upgrade_NN1 (_( escalation_NN1 )_) of_IO locks_NN2 ._. 
</s>
<s>
This_DD1 idea_NN1 is_VBZ to_TO separate_VVI the_AT bursts_NN2 of_IO the_AT PPBP_NP1 into_II long_JJ and_CC short_JJ bursts_NN2 ._. 
</s>
<s>
Enrollment_NN1 stage_NN1 :_: We_PPIS2 transformed_VVD the_AT gallery_NN1 images_NN2 into_II CLBP_NP1 transformed_VVD gallery_NN1 images_NN2 with_IW the_AT CLBP_NP1 codes_NN2 which_DDQ are_VBR obtained_VVN in_II the_AT training_NN1 stage_NN1 ._. 
</s>
<s>
Ideally_RR a_AT1 team_NN1 of_IO experts_NN2 is_VBZ required_VVN that_CST can_VM sufficiently_RR well_RR overlap_VVI with_IW one_PPX121 another_PPX122 in_II translating_VVG needs_NN2 and_CC requirements_NN2 ._. 
</s>
<s>
A_AT1 recent_JJ study_NN1 on_II the_AT prevalence_NN1 of_IO DDoS_NN2 in_II the_AT Internet_NP1 (_( Moore_NP1 et_RA21 al._RA22 ,_, 2006_MC )_) has_VHZ revealed_VVN that_CST a_AT1 significant_JJ number_NN1 of_IO DDoS_NP1 attacks_NN2 are_VBR being_VBG directed_VVN towards_II routers_NN2 ._. 
</s>
<s>
In_BCL21 order_BCL22 to_TO quantify_VVI the_AT performance_NN1 of_IO such_DA attacks_NN2 ,_, we_PPIS2 have_VH0 conducted_VVN the_AT following_JJ experiment_NN1 using_VVG the_AT Schonlau_NP1 et_RA21 al_RA22 ._. 
</s>
<s>
'_NULL s_ZZ1 dataset_NN1 ._. 
</s>
<s>
However_RR as_CSA we_PPIS2 were_VBDR looking_VVG for_IF an_AT1 efficient_JJ solution_NN1 ,_, we_PPIS2 further_RRR replaced_VVD it_PPH1 by_II our_APPGE HIST_JJ algorithm_NN1 which_DDQ provides_VVZ a_AT1 better_JJR solution_NN1 in_II31 terms_II32 of_II33 time_NNT1 complexity_NN1 and_CC provisioning_NN1 cost_NN1 ._. 
</s>
<s>
It_PPH1 is_VBZ important_JJ to_TO test_VVI forensic_JJ tools_NN2 with_IW base_NN1 installations_NN2 of_IO operating_NN1 systems_NN2 :_: characterizing_VVG the_AT tool_NN1 '_NULL s_ZZ1 behavior_NN1 on_II base_NN1 installs_VVZ helps_VVZ to_TO predict_VVI how_RRQ the_AT tool_NN1 will_VM behave_VVI on_II the_AT same_DA OS_NP2 files_NN2 when_CS they_PPHS2 are_VBR present_JJ on_II subject_NN1 media_NN ._. 
</s>
<s>
One_MC1 way_NN1 to_TO form_VVI such_DA a_AT1 network_NN1 is_VBZ to_TO use_VVI a_AT1 central_JJ coordinator_NN1 that_CST selects_VVZ which_DDQ nodes_NN2 are_VBR supernodes_NN2 and_CC assigns_VVZ them_PPHO2 responsibility_NN1 for_IF non-supernodes_NN2 ._. 
</s>
<s>
We_PPIS2 chose_VVD the_AT Enterprise_NN1 Application_NN1 Software_NN1 categorization_NN1 in_II the_AT Gartner_NN1 IT_NN1 Glossary13_FO as_CSA sufficiently_RR representative_JJ ._. 
</s>
<s>
One_MC1 consequence_NN1 of_IO this_DD1 process_NN1 is_VBZ that_CST recognised_VVD by_II Jedlitschka_NP1 and_CC Ciolkowski_NP1 &lsqb;_( 29_MC &rsqb;_) i.e._REX that_CST practitioners_NN2 abandon_VV0 some_DD of_IO the_AT projects_NN2 '_NULL '_NULL standard_JJ '_NULL processes_NN2 and_CC technologies_NN2 because_CS they_PPHS2 are_VBR perceived_VVN to_TO take_VVI more_DAR time_NNT1 than_CSN is_VBZ available_JJ to_II the_AT project_NN1 ._. 
</s>
<s>
Therefore_RR ,_, this_DD1 method_NN1 is_VBZ not_XX an_AT1 unsupervised_JJ method_NN1 to_TO obtain_VVI a_AT1 polygonal_JJ approximation_NN1 because_CS the_AT number_NN1 of_IO points_NN2 is_VBZ fixed_VVN ._. 
</s>
<s>
The_AT SDP_NP1 enables_VVZ applications_NN2 to_II interwork_NN1 with_IW each_PPX221 other_PPX222 ._. 
</s>
<s>
Roughly_RR speaking_VVG ,_, a_AT1 set_NN1 of_IO ground_NN1 atoms_NN2 is_VBZ externally_RR supported_VVN if_CS there_EX exists_VVZ a_AT1 ground_NN1 atom_NN1 in_II the_AT set_NN1 and_CC an_AT1 associated_JJ rule_NN1 that_CST supports_VVZ the_AT atom_NN1 (_( i.e._REX ,_, the_AT atom_NN1 is_VBZ the_AT head_NN1 of_IO the_AT ground_NN1 rule_NN1 )_) and_CC whose_DDQGE positive_JJ body_NN1 could_VM be_VBI satisfied_VVN by_II external_JJ ground_NN1 atoms_NN2 (_( i.e._REX ,_, ground_NN1 atoms_NN2 not_XX in_II this_DD1 set_NN1 )_) ._. 
</s>
<s>
In_II contrast_NN1 ,_, false_JJ negatives_NN2 are_VBR flows_NN2 or_CC packets_NN2 that_CST should_VM have_VHI been_VBN marked_VVN as_CSA malicious_JJ ,_, but_CCB were_VBDR flagged_VVN as_CSA non-malicious_JJ ._. 
</s>
<s>
The_AT Exchange_NN1 justified_VVN the_AT combination_NN1 by_II pointing_VVG out_RP that_CST some_DD customers_NN2 want_VV0 greater_JJR speed_NN1 and_CC choice_NN1 in_II order_NN1 execution_NN1 ,_, and_CC that_CST in_RR21 particular_RR22 institutional_JJ investors_NN2 want_VV0 to_TO purchase_VVI stocks_NN2 anonymously_RR ,_, something_PN1 hard_RR to_TO do_VDI with_IW the_AT physical_JJ floor_NN1 business_NN1 model_NN1 ._. 
</s>
<s>
For_REX21 instance_REX22 ,_, Glass_NP1 et_RA21 al._RA22 &lsqb;_( 12_MC &rsqb;_) studied_VVD a_AT1 5-year_JJ period_NN1 (_( 1995_MC to_II 1999_MC )_) across_II five_MC top_JJ journals_NN2 for_IF software_NN1 engineering_NN1 research_NN1 (_( Information_NN1 and_CC Software_NN1 Technology_NN1 ,_, Journal_NN1 of_IO Systems_NN2 and_CC Software_NN1 ,_, Software_NN1 Practice_NN1 and_CC Experience_NN1 ,_, IEEE_NP1 Software_NN1 ,_, ACM_NN1 Transactions_NN2 of_IO Software_NN1 Engineering_NN1 and_CC Methodology_NN1 ,_, and_CC IEEE_NP1 Transactions_NN2 on_II Software_NN1 Engineering_NN1 )_) ._. 
</s>
<s>
As_CSA before_RT ,_, each_DD1 threshold_NN1 has_VHZ been_VBN tuned_VVN so_BCL21 as_BCL22 to_TO limit_VVI the_AT false_JJ positive_JJ rate_NN1 to_II 5%_NNU ._. 
</s>
<s>
Clearly_RR ,_, the_AT packets_NN2 incur_VV0 delay_NN1 whenever_RRQV the_AT wireless_NN1 node_NN1 goes_VVZ to_II sleep_NN1 ._. 
</s>
<s>
This_DD1 includes_VVZ new_JJ quantitative_JJ synthesis_NN1 and_CC hypothesis_NN1 testing_VVG in_RP near_II real_JJ time_NNT1 ,_, as_CSA data_NN streaming_VVG from_II distributed_JJ instruments_NN2 ,_, to_TO transform_VVI raw_JJ data_NN into_II high_JJ level_JJ domain-dependent_JJ information_NN1 ._. 
</s>
<s>
Furthermore_RR ,_, the_AT fact_NN1 that_CST other_JJ results_NN2 from_II within_II this_DD1 sub-group_NN1 suggest_VV0 that_CST they_PPHS2 feel_VV0 the_AT least_RGT vulnerable_JJ ,_, there_EX is_VBZ an_AT1 interesting_JJ disparity_NN1 between_II their_APPGE confidence_NN1 and_CC their_APPGE capability_NN1 ._. 
</s>
<s>
However_RR ,_, its_APPGE use_NN1 forces_NN2 researchers_NN2 to_II better_JJR grasp_NN1 farmer_NN1 decision-making_NN1 processes_NN2 ,_, at_RR21 least_RR22 for_IF the_AT rules_NN2 concerned_JJ ._. 
</s>
<s>
To_TO test_VVI the_AT seal_NN1 detection_NN1 approach_NN1 ,_, we_PPIS2 have_VH0 constructed_VVN a_AT1 database_NN1 with_IW 370_MC documents_NN2 containing_VVG seals_NN2 of_IO English_JJ text_NN1 characters_NN2 ._. 
</s>
<s>
This_DD1 graph_NN1 helps_VVZ us_PPIO2 gain_VVI insights_NN2 on_II how_RGQ different_JJ placement_NN1 schemes_NN2 allocate_VV0 VD_NN1 requests_NN2 across_II data_NN centers_NN2 to_TO maximize_VVI ?_? 
</s>
<s>
Net_JJ Utility_NN1 ._. 
</s>
<s>
To_TO promote_VVI further_JJR inquiry_NN1 along_II these_DD2 lines_NN2 ,_, I_PPIS1 venture_VV0 some_DD conjectures_NN2 deserving_JJ of_IO further_JJR study_NN1 ._. 
</s>
<s>
The_AT larger_JJR of_IO these_DD2 subsets_NN2 ,_, called_VVN OWL_NN1 DL_MC ,_, restricts_VVZ OWL_NN1 in_II two_MC ways_NN2 ._. 
</s>
<s>
My_APPGE hope_NN1 is_VBZ that_CST this_DD1 technology_NN1 can_VM provide_VVI a_AT1 much-needed_JJ boost_NN1 in_II our_APPGE capability_NN1 to_TO address_VVI global_JJ environmental_JJ challenges_NN2 ._. 
</s>
<s>
We_PPIS2 have_VH0 combined_VVN a_AT1 number_NN1 of_IO techniques_NN2 developed_VVN over_II the_AT years_NNT2 for_IF different_JJ methods_NN2 and_CC demonstrated_VVD that_CST the_AT resulting_JJ method_NN1 is_VBZ accurate_JJ and_CC quite_RG viable_JJ ._. 
</s>
<s>
Therefore_RR ,_, when_CS releasing_VVG a_AT1 real_JJ lock_NN1 on_II an_AT1 item_NN1 that_CST has_VHZ locked_VVN children_NN2 ,_, we_PPIS2 can_VM not_XX simply_RR release_VVI the_AT lock_NN1 ._. 
</s>
<s>
A_AT1 valid_JJ decision_NN1 sequence_NN1 begins_VVZ with_IW either_RR a_AT1 delegation_NN1 or_CC refinement_NN1 decisiond1∈_FO (_( DSRS_NP1 )_) and_CC alternates_VVZ between_II zero_MC or_CC more_DAR delegation_NN1 and_CC refinement_NN1 sequences_NN2 ._. 
</s>
<s>
This_DD1 loading_NN1 path_NN1 is_VBZ discretized_VVN into_II n1n1_FO loading_NN1 steps_NN2 ,_, and_CC we_PPIS2 denote_VV0 with_IW TkTk_NP1 ,_, fkfk_NNU and_CC @k_FO the_AT applied_JJ tractions_NN2 ,_, body_NN1 forces_NN2 and_CC prescribed_JJ displacements_NN2 at_II step_NN1 k_ZZ1 ._. 
</s>
<s>
An_AT1 open_JJ space_NN1 is_VBZ a_AT1 space_NN1 in_II which_DDQ traveling_NN1 from_II any_DD distinct_JJ recorded_JJ location_NN1 to_II any_DD other_JJ distinct_JJ recorded_JJ location_NN1 can_VM be_VBI done_VDN without_IW having_VHG to_TO pass_VVI through_II a_AT1 3rdspecific_FO recorded_JJ location_NN1 (_( e.g._REX ,_, the_AT only_JJ way_NN1 from_II A_ZZ1 to_II B_ZZ1 is_VBZ through_II a_AT1 door_NN1 that_CST records_VVZ entries_NN2 )_) ._. 
</s>
<s>
For_REX21 instance_REX22 ,_, during_II the_AT British_JJ royal_JJ wedding_NN1 in_II 2011_MC ,_, tweets_VVZ during_II the_AT event_NN1 exceeded_VVD 1_MC1 million_NNO ._. 
</s>
<s>
Unfortunately_RR it_PPH1 has_VHZ been_VBN shown_VVN that_CST this_DD1 popular_JJ prediction_NN1 accuracy_NN1 statistic_NN1 is_VBZ flawed_JJ in_CS21 that_CS22 it_PPH1 is_VBZ a_AT1 biased_JJ estimator_NN1 of_IO central_JJ tendency_NN1 of_IO the_AT residuals_NN2 of_IO a_AT1 prediction_NN1 system_NN1 because_CS it_PPH1 is_VBZ an_AT1 asymmetric_JJ measure_NN1 ._. 
</s>
<s>
For_IF each_DD1 topic_NN1 identified_VVN in_II the_AT requirements_NN2 analysis_NN1 ,_, a_AT1 scenario_NN1 element_NN1 was_VBDZ created_VVN that_CST requires_VVZ the_AT player_NN1 to_TO do_VDI something_PN1 that_CST will_VM convey_VVI the_AT concept_NN1 to_TO be_VBI learned_VVN ._. 
</s>
<s>
These_DD2 studies_NN2 compared_VVD disposable_JJ beverage_NN1 cups_NN2 made_VVN from_II different_JJ types_NN2 of_IO material_NN1 such_II21 as_II22 paper_NN1 ,_, petro-plastics_NN2 ,_, and_CC bioplastics_NN2 ._. 
</s>
<s>
The_AT community_NN1 consensus_NN1 was_VBDZ that_CST the_AT terms_NN2 generation_NN1 ,_, use_VV0 ,_, derivation_NN1 ,_, and_CC version_NN1 should_VM be_VBI adopted_VVN for_IF these_DD2 notions_NN2 ,_, respectively_RR ._. 
</s>
<s>
Then_RT ,_, when_CS a_AT1 test_NN1 case_NN1 in_II the_AT prioritized_JJ order_NN1 or_CC the_AT reduced_JJ suite_NN1 is_VBZ executed_VVN ,_, the_AT state_NN1 of_IO the_AT application_NN1 is_VBZ reset_VVN to_II the_AT stored_JJ state_NN1 of_IO the_AT test_NN1 case_NN1 from_II the_AT original_JJ suite_NN1 execution_NN1 ._. 
</s>
<s>
Similarly_RR ,_, Baskerville_NP1 (_( 1993_MC )_) presented_VVD the_AT need_NN1 for_IF the_AT future_JJ generation_NN1 of_IO security_NN1 tools_NN2 that_CST provide_VV0 automated_JJ features_NN2 ._. 
</s>
<s>
For_IF our_APPGE application_NN1 ,_, we_PPIS2 definitely_RR want_VV0 to_TO consider_VVI ?_? d1_FO ?_? and_CC ?_? d2similar_FO ,_, which_DDQ requires_VVZ ?_? ?_? ≥_FO 0.30_MC ._. 
</s>
<s>
The_AT questionnaire_NN1 was_VBDZ pre-tested_JJ by_II selected_JJ members_NN2 in_II31 charge_II32 of_II33 IS_VBZ from_II different_JJ universities_NN2 in_II the_AT UAE_NP1 ._. 
</s>
<s>
This_DD1 stems_VVZ from_II the_AT desire_NN1 to_TO be_VBI able_JK to_TO manage_VVI RDF_NP1 data_NN in_II a_AT1 secure_JJ ,_, scalable_JJ ,_, and_CC highly_RR available_JJ environment_NN1 ._. 
</s>
<s>
This_DD1 could_VM be_VBI solved_VVN by_II either_RR tweaking_VVG the_AT B+Tree_FO parameters_NN2 for_IF this_DD1 large_JJ amount_NN1 of_IO versions_NN2 ,_, reducing_VVG storage_NN1 requirements_NN2 for_IF each_DD1 value_NN1 ,_, or_CC by_II dynamically_RR creating_VVG a_AT1 new_JJ snapshot_NN1 ._. 
</s>
<s>
It_PPH1 is_VBZ worth_II mentioning_VVG that_CST the_AT fact_NN1 that_CST RSTP_NP1 processes_VVZ the_AT resource_NN1 failures_NN2 immediately_RR while_CS BLSTR_NP1 processes_VVZ the_AT resource_NN1 failures_NN2 at_II the_AT time_NNT1 at_II which_DDQ they_PPHS2 have_VH0 been_VBN received_VVN by_II all_DB the_AT network_NN1 bridges_NN2 (_( with_IW a_AT1 tolerance_NN1 of_IO ?_? 2TS_FO )_) requires_VVZ the_AT adoption_NN1 of_IO a_AT1 specific_JJ technical_JJ solution_NN1 ,_, not_XX covered_VVN in_II this_DD1 presentation_NN1 ,_, to_TO control_VVI the_AT transition_NN1 from_II state_NN1 ?_? 
</s>
<s>
RSTP_NP1 ?_? to_II stateNO_NN1 ._. 
</s>
<s>
Perry_NP1 and_CC Millington_NP1 (_( 2008_MC )_) distinguish_VV0 the_AT complementary_JJ approaches_NN2 of_IO predictive_JJ and_CC exploratory_JJ spatial_JJ modelling_NN1 of_IO succession-disturbance_JJ dynamics_NN in_II forest_NN1 ecosystems_NN2 ._. 
</s>
<s>
The_AT BGP_NP1 update_NN1 streams_NN2 were_VBDR collected_VVN from_II the_AT at_II the_AT University_NN1 of_IO Oregon_NP1 ._. 
</s>
<s>
All_DB stakeholders_NN2 are_VBR also_RR part_NN1 of_IO a_AT1 larger_JJR community_NN1 (_( e.g._REX ,_, the_AT general_JJ public_NN1 in_II the_AT city_NN1 of_IO Merida_NP1 )_) that_CST affects_VVZ and_CC is_VBZ affected_VVN by_II these_DD2 interactions_NN2 ._. 
</s>
<s>
There_EX are_VBR possible_JJ ways_NN2 to_TO marginally_RR improve_VVI efficiency_NN1 of_IO the_AT proposed_JJ ACO_NN1 procedure_NN1 ._. 
</s>
<s>
If_CS successful_JJ ,_, this_DD1 proves_VVZ that_CST the_AT tag_NN1 knows_VVZ ?_? 
</s>
<s>
KTID_NP1 ?_? and_CC the_AT tag_NN1 is_VBZ authenticated_VVN by_II the_AT server_NN1 ._. 
</s>
<s>
Comparing_VVG two_MC incapacitation_NN1 periods_NN2 (_( 6_MC and_CC 24_MC months_NNT2 )_) ._. 
</s>
<s>
The_AT two_MC cases_NN2 to_TO consider_VVI based_VVN on_II &lsqb;_( 4_MC &rsqb;_) namely_REX those_DD2 when_RRQ ei_NNU =_FO 0_MC and_CC ei_NNU =_FO 1_MC1 ,_, correspond_VV0 to_II the_AT bounding_JJ polynomials_NN2 for_IF PN(s)_NP1 ._. 
</s>
<s>
These_DD2 low-volume_JJ transactors_NN2 never_RR incur_VV0 finance_NN1 charges_NN2 ,_, and_CC for_IF the_AT issuer_NN1 serving_VVG them_PPHO2 represent_VV0 little_DA1 more_DAR than_CSN supporting_VVG a_AT1 collection_NN1 of_IO perpetual_JJ ,_, zero-interest_JJ loans_NN2 with_IW monthly_JJ mailings_NN2 ._. 
</s>
<s>
Whereas_CS the_AT W3C_FO Recommendations_NN2 and_CC Notes_NN2 focus_VV0 on_II the_AT technical_JJ specification_NN1 of_IO prov_NN1 ,_, and_CC publications_NN2 such_II21 as_II22 &lsqb;_( 22_MC &rsqb;_) focus_VV0 on_II the_AT use_NN1 and_CC practical_JJ deployment_NN1 of_IO prov_NN1 ,_, this_DD1 article_NN1 ,_, in_II contrast_NN1 ,_, is_VBZ concerned_JJ with_IW the_AT rationale_NN1 for_IF prov_NN1 ._. 
</s>
<s>
In_II other_JJ words_NN2 ,_, if_CS the_AT current_JJ escapement_NN1 is_VBZ much_RR lower_JJR than_CSN the_AT pristine_JJ one_PN1 (_( current/pristine_JJ ratio_NN1 <<0.4_FO )_) ,_, an_AT1 EMP_NN1 might_VM have_VHI not_XX only_RR to_TO close_VVI the_AT fishery_NN1 ,_, but_CCB also_RR to_TO remove_VVI obstacles_NN2 to_II migration_NN1 to_TO achieve_VVI the_AT potential_JJ escapement_NN1 ._. 
</s>
<s>
The_AT single-axis_JJ topology_NN1 and_CC the_AT multi-axis_JJ topology_NN1 results_NN2 are_VBR separately_RR analyzed_VVN below_RL ._. 
</s>
<s>
We_PPIS2 therefore_RR rely_VV0 on_II complementary_JJ resources_NN2 such_II21 as_II22 SP_NN1 and_CC transmembrane_NN1 protein_NN1 data_NN to_TO develop_VVI the_AT PTPF_NP1 and_CC to_TO study_VVI SRP-binding_JJ features_NN2 ._. 
</s>
<s>
Also_RR ,_, the_AT accuracy_NN1 of_IO the_AT derived_JJ topographical_JJ parameters_NN2 was_VBDZ identified_VVN to_TO strongly_RR depend_VVI on_II (_( 1_MC1 )_) the_AT resolution_NN1 of_IO the_AT used_JJ DEM_FW (_( Vaze_NP1 et_NP1 ?_? al._NNU ,_, 2010_MC )_) and_CC (_( 2_MC )_) the_AT errors_NN2 and_CC uncertainty_NN1 of_IO the_AT DEM_FW (_( Raaflaub_NP1 and_CC Collins_NP1 ,_, 2006_MC ,_, Gericke_NP1 and_CC Venohr_NP1 ,_, 2012_MC )_) ._. 
</s>
<s>
While_CS the_AT manual_JJ formulation_NN1 of_IO the_AT archetypes_NN2 provides_VVZ us_PPIO2 an_AT1 initial_JJ framework_NN1 of_IO discernment_NN1 ,_, in_II the_AT next_MD phase_NN1 we_PPIS2 proceeded_VVD to_II standard_NN1 ,_, automatic_JJ cluster_NN1 analysis_NN1 &lsqb;_( 19_MC &rsqb;_) of_IO the_AT case-study_JJ data_NN ._. 
</s>
<s>
Regarding_II the_AT identity-based_NN1 and_CC hybrid_JJ approach_NN1 ,_, the_AT assumption_NN1 of_IO single_JJ concept_NN1 correspondence_NN1 is_VBZ made_VVN ,_, in_II accordance_NN1 to_II the_AT notion_NN1 of_IO unique_JJ "_" identity_NN1 "_" ,_, as_CSA initially_RR introduced_VVN for_IF the_AT first_MD approach_NN1 in_II &lsqb;_( 13_MC &rsqb;_) ._. 
</s>
<s>
Despite_II being_VBG outperformed_VVN by_II SVM_NP1 classifiers_NN2 ,_, the_AT FOIL_NN1 based_VVN classifiers_NN2 have_VH0 a_AT1 number_NN1 of_IO significant_JJ advantages_NN2 ._. 
</s>
<s>
The_AT proposed_JJ algorithm_NN1 is_VBZ composed_VVN of_IO three_MC sub-algorithms_NN2 ,_, which_DDQ deal_VV0 with_IW the_AT refinement_NN1 of_IO end_NN1 ,_, normal_JJ and_CC junction_NN1 triangles_NN2 ,_, the_AT three_MC main_JJ components_NN2 of_IO the_AT CDT_NP1 method_NN1 ._. 
</s>
<s>
The_AT Global_JJ Internet_NN1 is_VBZ a_AT1 collection_NN1 of_IO realms_NN2 &lsqb;_( 124_MC &rsqb;_) ,_, &lsqb;_( 125_MC &rsqb;_) and&lsqb;126&rsqb;_FO of_IO disparate_JJ technologies_NN2 &lsqb;_( 102_MC &rsqb;_) ._. 
</s>
<s>
The_AT number_NN1 of_IO times_NNT2 that_CST the_AT over-sampling_NN1 procedure_NN1 is_VBZ applied_VVN is_VBZ the_AT number_NN1 of_IO classes_NN2 of_IO the_AT problem_NN1 ._. 
</s>
<s>
The_AT FAML_NN1 agent-external_JJ system_NN1 diagram_NN1 describes_VVZ an_AT1 agent_NN1 system_NN1 ._. 
</s>
<s>
Apparently_RR manually_RR enforcing_VVG a_AT1 threshold_NN1 and_CC assuming_VVG computation_NN1 correct_JJ if_CS differences_NN2 are_VBR less_DAR than_CSN that_DD1 threshold_NN1 is_VBZ unreliable_JJ ._. 
</s>
<s>
Analysts_NN2 might_VM err_VVI when_CS capturing_VVG an_AT1 argument_NN1 ._. 
</s>
<s>
Larger_JJR rallows_NN2 for_IF tolerating_VVG more_RGR simultaneous_JJ node_NN1 failures_NN2 ,_, albeit_CS at_II higher_JJR cost_NN1 ._. 
</s>
<s>
To_TO remove_VVI the_AT boundary_NN1 conditions_NN2 we_PPIS2 consider_VV0 this_DD1 model_NN1 as_CSA toroidal_JJ shape_NN1 where_CS each_DD1 cell_NN1 has_VHZ an_AT1 identical_JJ neighborhood_NN1 ._. 
</s>
<s>
The_AT overall_JJ procedure_NN1 is_VBZ summarized_VVN in_II Fig._NN1 2_MC in_II which_DDQ the_AT label_NN1 for_IF each_DD1 class_NN1 is_VBZ assumed_VVN to_TO be_VBI +1_MC or_CC ?_? 1_MC1 ._. 
</s>
<s>
A_AT1 promising_JJ modelling_NN1 strategy_NN1 is_VBZ to_TO describe_VVI each_DD1 functional_JJ component_NN1 with_IW a_AT1 neuronal_JJ network_NN1 or_CC mean-field_JJ model_NN1 ,_, and_CC then_RT have_VH0 them_PPHO2 interact_VV0 according_II21 to_II22 empirically_RR determined_JJ coupling_NN1 ,_, thus_RR combining_VVG the_AT forward_NN1 and_CC backward_JJ approaches_NN2 &lsqb;_( 1_MC1 &rsqb;_) ._. 
</s>
<s>
The_AT context_NN1 in&lsqb;6&rsqb;is_FO significantly_RR different_JJ from_II here_RL ,_, so_RR deducing_VVG a_AT1 specific_JJ form_NN1 ,_, from_II this_DD1 result_NN1 ,_, for_IF the_AT weight_NN1 of_IO the_AT tail_NN1 which_DDQ applies_VVZ in_II the_AT present_JJ case_NN1 would_VM be_VBI difficult_JJ ._. 
</s>
<s>
This_DD1 file_NN1 contained_VVD a_AT1 hierarchical_JJ representation_NN1 of_IO the_AT document_NN1 structure_NN1 :_: abstract_JJ ,_, sections_NN2 and_CC subsections_NN2 ,_, list_NN1 of_IO references_NN2 ,_, footnotes_NN2 ,_, appendices_NN2 ,_, tables_NN2 and_CC figures_NN2 ._. 
</s>
<s>
On_II the_AT contrary_NN1 ,_, the_AT state_NN1 of_IO the_AT AP_NP1 is_VBZ dictated_VVN by_II the_AT sleep/awake_VV0 duration_NN1 of_IO the_AT wireless_NN1 nodes_NN2 in_II the_AT system_NN1 ._. 
</s>
<s>
Ontologies_NN2 have_VH0 more_DAR flexibility_NN1 than_CSN set_JJ standards_NN2 ,_, they_PPHS2 simplify_VV0 policy_NN1 specification_NN1 ,_, and_CC they_PPHS2 enable_VV0 more_DAR information_NN1 to_TO be_VBI specified_VVN to_TO control_VVI privacy_NN1 during_II trust_NN1 negotiation_NN1 ._. 
</s>
<s>
Thus_RR hypothesis_NN1 3a_FO was_VBDZ not_XX supported_VVN for_IF reactive_JJ market_NN1 orientation_NN1 ._. 
</s>
<s>
The_AT terms_NN2 BuTDBu_NN1 and_CC H2BuTDBu_FO are_VBR composed_VVN of_IO polynomials_NN2 on_II triangular_JJ and_CC rectangular_JJ elements_NN2 and_CC are_VBR not_XX discontinuous_JJ inside_II an_AT1 element_NN1 ,_, whereas_CS the_AT term_NN1 HBuTDBu_NN1 is_VBZ composed_VVN of_IO polynomials_NN2 multiplied_VVN by_II the_AT Heaviside_NN1 function_NN1 with_IW a_AT1 discontinuity_NN1 across_II the_AT crack_NN1 ._. 
</s>
<s>
Moreover_RR ,_, a_AT1 review_NN1 of_IO recent_JJ studies_NN2 of_IO Knowledge_NN1 Management_NN1 in_II31 support_II32 of_II33 Enterprise_NN1 Systems_NN2 ,_, suggests_VVZ other_JJ limitations_NN2 of_IO past_JJ research_NN1 in_II the_AT area_NN1 ._. 
</s>
<s>
First_MD a_AT1 filter_NN1 approach_NN1 is_VBZ used_VVN to_TO calculate_VVI the_AT information_NN1 gain_NN1 ratio_NN1 of_IO each_DD1 feature_NN1 individually_RR ._. 
</s>
<s>
It_PPH1 is_VBZ based_VVN on_II a_AT1 static_JJ rule_NN1 set_VVD :_: the_AT system_NN1 can_VM not_XX adapt_VVI the_AT filter_NN1 to_TO identify_VVI emerging_JJ spam_NN1 characteristics_NN2 ._. 
</s>
<s>
The_AT idea_NN1 of_IO using_VVG these_DD2 data_NN is_VBZ to_TO be_VBI able_JK to_TO compare_VVI the_AT results_NN2 of_IO our_APPGE multidimensional_JJ PDF_NP1 index_NN1 with_IW already_RR published_VVN results_NN2 (_( Errasti_NP1 et_NP1 ?_? al._NNU ,_, 2013_MC )_) that_CST used_VVD alternative_JJ techniques_NN2 in_II31 terms_II32 of_II33 the_AT skill_NN1 score_NN1 of_IO the_AT models_NN2 ._. 
</s>
<s>
The_AT process_NN1 of_IO obtaining_VVG and_CC using_VVG the_AT OTP_NN1 was_VBDZ very_RG straightforward_JJ ._. 
</s>
<s>
While_CS many_DA2 have_VH0 called_VVN for_IF investigations_NN2 into_II the_AT capability_NN1 development_NN1 process_NN1 (_( &lsqb;_( Oliver_NP1 ,_, 1997_MC &rsqb;_) ,_, &lsqb;_( Melville_NP1 et_RA21 al._RA22 ,_, 2004_MC &rsqb;_) and_CC &lsqb;_( Wade_NP1 and_CC Hulland_NP1 ,_, 2004_MC &rsqb;_) )_) ,_, this_DD1 is_VBZ one_MC1 of_IO the_AT first_MD studies_NN2 to_TO highlight_VVI the_AT critical_JJ role_NN1 of_IO non-executive_JJ boundary_NN1 spanners_NN2 in_II EMP_NP1 development_NN1 ._. 
</s>
<s>
Explicit_JJ semantics_NN1 may_VM be_VBI used_VVN to_TO represent_VVI knowledge_NN1 in_II the_AT Grid_NN1 environment_NN1 ,_, the_AT source_NN1 of_IO which_DDQ could_VM come_VVI from_II each_DD1 of_IO the_AT three_MC tiers_NN2 (_( i.e._REX application_NN1 ,_, middleware_NN1 ,_, fabric_NN1 )_) of_IO OGSA_NN1 ._. 
</s>
<s>
For_REX21 example_REX22 ,_, Java_NP1 '_NULL s_ZZ1 SQLPermission_NN1 allows_VVZ a_AT1 program_NN1 to_TO set_VVI the_AT logging_NN1 stream_NN1 that_CST may_VM contain_VVI private_JJ SQL_NN1 data_NN ;_; it_PPH1 is_VBZ checked_VVN before_II setLogWriter_NN1 methods_NN2 in_II several_DA2 classes_NN2 ._. 
</s>
<s>
Each_DD1 one_PN1 of_IO these_DD2 tables_NN2 contains_VVZ a_AT1 column_NN1 where_RRQ geometries_NN2 are_VBR stored_VVN in_II binary_JJ format_NN1 (_( WKB_NP1 )_) and_CC an_AT1 index_NN1 has_VHZ been_VBN built_VVN on_II that_DD1 column_NN1 ._. 
</s>
<s>
In_II the_AT fullness_NN1 of_IO time_NNT1 we_PPIS2 hope_VV0 to_TO provide_VVI an_AT1 automatable_JJ method_NN1 to_TO facilitate_VVI requirements_NN2 engineering_NN1 ._. 
</s>
<s>
For_IF each_DD1 interior_JJ face_NN1 γ_NULL f=r_FO ,_, we_PPIS2 writenandnrfor_NN1 the_AT unit_NN1 outer_JJ normal_JJ toandr_NN1 ,_, respectively_RR ._. 
</s>
<s>
The_AT analytic_JJ and_CC simulation_NN1 results_NN2 are_VBR summarized_VVN inTable_JJ 6_MC ,_, where_CS it_PPH1 is_VBZ seen_VVN that_CST the_AT entries_NN2 offall_VV0 within_II the_AT 95%_NNU confidence_NN1 intervals_NN2 for_IF ._. 
</s>
<s>
Next_MD ,_, each_DD1 car_NN1 probabilistically_RR disseminates_VVZ the_AT pieces_NN2 using_VVG an_AT1 epidemic_NN1 (_( "_" gossip_NN1 "_" )_) scheme_NN1 (_( we_PPIS2 later_RRR review_VV0 the_AT effectiveness_NN1 of_IO epidemic_NN1 dissemination_NN1 in_II the_AT vehicular_JJ context_NN1 )_) ._. 
</s>
<s>
The_AT best_JJT search_NN1 gave_VVD a_AT1 considerably_RR better_RRR (_( smaller_JJR )_) result_NN1 than_CSN the_AT average_NN1 ,_, at_II the_AT expense_NN1 of_IO more_RGR computational_JJ effort_NN1 as_CSA shown_VVN by_II the_AT numbers_NN2 of_IO jobs_NN2 and_CC batches_NN2 of_IO jobs_NN2 ._. 
</s>
<s>
Second_MD ,_, we_PPIS2 choose_VV0 the_AT modeled_JJ membrane_NN1 shear_VV0 modulus_NN1 0_MC ,_, and_CC area_NN1 and_CC volume_NN1 constraint_NN1 coefficients_NN2 (_( Eqs_NN2 ._. 
</s>
<s>
Otherwise_RR ,_, it_PPH1 is_VBZ also_RR valuable_JJ to_TO consider_VVI this_DD1 process_NN1 as_CSA one_PN1 that_CST evolves_VVZ and_CC is_VBZ emergent_JJ as_CSA learner_NN1 feedback_NN1 from_II social_JJ media_NN outlets_NN2 is_VBZ incorporated_VVN into_II the_AT total_JJ transmedia_NN1 learning_NN1 design_NN1 ._. 
</s>
<s>
Also_RR ,_, this_DD1 approach_NN1 involves_VVZ higher_JJR computational_JJ cost_NN1 because_II21 of_II22 the_AT need_NN1 to_TO perform_VVI the_AT identification_NN1 multiple_JJ times_NNT2 ._. 
</s>
<s>
We_PPIS2 have_VH0 also_RR defined_VVN a_AT1 criteria_NN2 for_IF quality_NN1 assessment_NN1 of_IO included_JJ studies_NN2 based_VVN on_II the_AT aims_NN2 and_CC research_NN1 questions_NN2 of_IO our_APPGE study_NN1 ._. 
</s>
<s>
If_CS validation_NN1 is_VBZ not_XX properly_RR performed_VVN ,_, a_AT1 traffic_NN1 simulation_NN1 model_NN1 may_VM not_XX provide_VVI accurate_JJ results_NN2 and_CC should_VM not_XX be_VBI used_VVN to_TO make_VVI important_JJ decisions_NN2 with_IW financial_JJ ,_, environmental_JJ and_CC social_JJ impacts_NN2 ._. 
</s>
<s>
From_II Table_NN1 5_MC ,_, it_PPH1 can_VM be_VBI noted_VVN that_CST IFSPA-IVPR_NP1 has_VHZ a_AT1 far_RG lower_JJR mean_JJ computational_JJ time_NNT1 and_CC standard_JJ deviation_NN1 than_CSN the_AT ones_NN2 produced_VVN by_II the_AT original_JJ IVPR_NN1 ._. 
</s>
<s>
To_TO better_RRR illustrate_VVI the_AT workings_NN2 of_IO subdimensional_JJ expansion_NN1 ,_, we_PPIS2 present_VV0 an_AT1 example_NN1 for_IF multirobot_NN1 path_NN1 planning_VVG on_II graphs_NN2 ._. 
</s>
<s>
A_AT1 centralized_JJ manager_NN1 identifies_VVZ the_AT available_JJ resources_NN2 without_IW any_DD interaction_NN1 with_IW the_AT potential_JJ lenders_NN2 ._. 
</s>
<s>
In_II the_AT first_MD three_MC examples_NN2 we_PPIS2 solve_VV0 each_DD1 problem_NN1 using_VVG a_AT1 fixed_JJ fine_JJ grid_NN1 several_DA2 times_NNT2 ._. 
</s>
<s>
Table_NN1 1_MC1 presents_VVZ the_AT average_JJ number_NN1 of_IO edges_NN2 per_II instance_NN1 graph_NN1 or_CC tree_NN1 for_IF the_AT datasets_NN2 used_VVN in_II the_AT classification_NN1 experiments_NN2 ._. 
</s>
<s>
However_RR ,_, once_RR a_AT1 promising_JJ outsourcing_JJ candidate_NN1 project_NN1 has_VHZ been_VBN identified_VVN ,_, the_AT most_RGT dominant_JJ criteria_NN2 that_CST they_PPHS2 use_VV0 to_TO choose_VVI a_AT1 vendor_NN1 is_VBZ the_AT relative_JJ cost_NN1 advantage_NN1 that_CST a_AT1 vendor_NN1 offers_NN2 ,_, as_CSA our_APPGE empirical_JJ results_NN2 suggest_VV0 ._. 
</s>
<s>
There_EX are_VBR four_MC types_NN2 of_IO transfers_NN2 used_VVN by_II USB_NP1 devices_NN2 :_: Bulk_VV0 ,_, Control_NN1 ,_, Interrupt_VV0 and_CC Isochronous_NP1 ._. 
</s>
<s>
Thecontextis_NN1 that_CST the_AT functionality_NN1 under_II test_NN1 is_VBZ captured_VVN as_II a_AT1 UCM_JJ path_NN1 that_CST contains_VVZ multiple_JJ causally_RR linked_VVN dynamic_JJ stubs_NN2 ._. 
</s>
<s>
The_AT primary_JJ focus_NN1 of_IO the_AT previous_JJ research_NN1 has_VHZ been_VBN on_II the_AT formation_NN1 of_IO behavioral_JJ intention_NN1 to_TO measure_VVI the_AT actual_JJ information_NN1 technology_NN1 (_( IT_NN1 )_) behaviors_NN2 almost_RR to_II the_AT exclusion_NN1 of_IO other_JJ factors_NN2 that_CST would_VM affect_VVI the_AT actual_JJ behavior_NN1 of_IO the_AT respondent_NN1 (_( Limayem_NP1 et_RA21 al._RA22 ,_, 2007_MC )_) ._. 
</s>
<s>
For_IF Fig._NN1 5(d)_FO we_PPIS2 get_VV0 (_( fc_NNU )_) avg=1.08_FO and_CC therefore_RR we_PPIS2 get_VV0 γ_NULL ≈_NULL 1_MC1 ._. 
</s>
<s>
The_AT RSV_NP1 values_NN2 were_VBDR close_JJ to_II 1_MC1 when_CS the_AT SCM_NP1 was_VBDZ used_VVN ._. 
</s>
<s>
They_PPHS2 found_VVD that_CST the_AT total_JJ amount_NN1 of_IO nitrogen_NN1 was_VBDZ higher_JJR in_II impervious_JJ urban_JJ areas_NN2 than_CSN previous_JJ ones_NN2 ._. 
</s>
<s>
MUSIC_NN1 was_VBDZ insensitive_JJ to_II some_DD of_IO the_AT dry_JJ weather_NN1 associated_JJ parameters_NN2 for_IF all_DB catchments_NN2 (_( coeff_VV0 ,_, sq_NNU ,_, SIni_NP1 and_CC gw_NNU )_) ._. 
</s>
<s>
The_AT accuracy_NN1 of_IO neurosurgery_NN1 is_VBZ not_XX better_RRR than_CSN 1mm&lsqb;1&rsqb;_FO ._. 
</s>
<s>
While_CS the_AT latter_DA has_VHZ been_VBN addressed_VVN using_VVG the_AT all-node_NN1 and_CC all-edge_JJ coverage_NN1 criteria_NN2 proposed_VVN in_II prior_JJ work_NN1 &lsqb;_( 37_MC &rsqb;_) ,_, this_DD1 study_NN1 has_VHZ introduced_VVN new_JJ criteria_NN2 aimed_VVN at_II communication-based_JJ coverage_NN1 (_( ,_, ,_, and_CC n-step-message-transfer_NN1 )_) ,_, and_CC defined_VVD the_AT hierarchy_NN1 and_CC intersection_NN1 of_IO their_APPGE coverage_NN1 domains_NN2 ._. 
</s>
<s>
But_CCB it_PPH1 is_VBZ difficult_JJ to_TO get_VVI a_AT1 clear_JJ view_NN1 ,_, from_II the_AT literature_NN1 ,_, of_IO what_DDQ could_VM be_VBI an_AT1 appropriate_JJ architecture_NN1 for_IF feature_NN1 fusion_NN1 in_II the_AT context_NN1 of_IO a_AT1 large_JJ size_NN1 feature_NN1 set_NN1 ._. 
</s>
<s>
It_PPH1 is_VBZ not_XX the_AT intention_NN1 to_TO require_VVI the_AT device_NN1 user_NN1 to_TO respond_VVI to_II an_AT1 explicit_JJ request_NN1 for_IF voice_NN1 input_NN1 ._. 
</s>
<s>
Thus_RR ,_, the_AT Welch/Brown-Forsythe_NN1 Robust_JJ Test_NN1 of_IO Equality_NN1 of_IO Means_NN is_VBZ used_VVN ,_, and_CC the_AT asymptotically_RR distributed_VVN F_ZZ1 statistic_NN1 reported_VVD ._. 
</s>
<s>
As_II the_AT common_JJ and_CC variation_NN1 code_NN1 files_NN2 are_VBR reused_VVN across_II products_NN2 ,_, they_PPHS2 go_VV0 through_II iterative_JJ cycles_NN2 of_IO testing_NN1 ,_, operation_NN1 and_CC maintenance_NN1 that_CST over_II time_NNT1 identify_VV0 and_CC remove_VV0 many_DA2 of_IO the_AT bugs_NN2 that_CST can_VM lead_VVI to_II failures_NN2 ._. 
</s>
<s>
Following_VVG TMDL_NP1 development_NN1 ,_, a_AT1 basin_NN1 planning_NN1 process_NN1 was_VBDZ initiated_VVN by_II the_AT Vermont_NP1 Agency_NN1 of_IO Natural_JJ Resources_NN2 that_CST involved_VVD a_AT1 substantial_JJ public_JJ involvement_NN1 component_NN1 to_TO incorporate_VVI public_JJ feedback_NN1 on_II proposed_JJ basin_NN1 plans_NN2 ._. 
</s>
<s>
Regardless_RR of_IO the_AT sophistication_NN1 of_IO the_AT statistical_JJ techniques_NN2 ,_, causal_JJ inferences_NN2 must_VM be_VBI treated_VVN with_IW caution_NN1 when_CS using_VVG non-experimental_JJ designs_NN2 ._. 
</s>
<s>
The_AT updated_JJ information_NN1 is_VBZ then_RT available_JJ in_II the_AT next_MD iteration_NN1 of_IO the_AT Evaluation_NN1 Stack_NN1 ._. 
</s>
<s>
Case_NN1 @=@0.15_FO fell_VVD in_II Regime_NN1 1_MC1 hence_RR (_( Table_NN1 2_MC )_) ,_, did_VDD not_XX present_VVI an_AT1 evolution_NN1 of_IO solitary_JJ waves_NN2 ._. 
</s>
<s>
Hence_RR ,_, assignment_NN1 is_VBZ the_AT process_NN1 where_CS an_AT1 organization_NN1 gives_VVZ an_AT1 AS_II the_AT right_JJ to_TO originate_VVI a_AT1 set_NN1 of_IO addresses_NN2 ._. 
</s>
<s>
Therefore_RR only_JJ subdivision_NN1 is_VBZ required_VVN ,_, with_IW the_AT sets_NN2 being_VBG restored_VVN on_II backtracking_VVG ._. 
</s>
<s>
In_BCL21 order_BCL22 to_TO compute_VVI the_AT cluster_NN1 attracted_VVN by_II a_AT1 given_JJ data_NN point_VV0 one_PN1 should_VM compute_VVI the_AT whole_JJ affinity_NN1 matrix_NN1 ._. 
</s>
<s>
There_EX was_VBDZ a_AT1 large_JJ difference_NN1 in_II price_NN1 between_II providers_NN2 for_IF the_AT same_DA level_NN1 of_IO performance_NN1 for_IF a_AT1 given_JJ service_NN1 ._. 
</s>
<s>
The_AT Semantic_JJ Web_NN1 has_VHZ solved_VVN different_JJ problems_NN2 ._. 
</s>
<s>
Configurations_NN2 for_IF which_DDQ BB_NP1 is_VBZ coercive_JJ are_VBR said_VVN to_TO be_VBI elastically_RR linearly_RR stable_JJ ._. 
</s>
<s>
Even_RR simple_JJ problems_NN2 like_II determining_VVG whether_CSW two_MC users_NN2 with_IW the_AT same_DA first_MD and_CC last_MD names_NN2 are_VBR the_AT same_DA person_NN1 across_II applications_NN2 involves_VVZ this_DD1 kind_NN1 of_IO analysis_NN1 ._. 
</s>
<s>
We_PPIS2 did_VDD not_XX experiment_VVI with_IW advisors_NN2 ._. 
</s>
<s>
Word_NN1 clouds_NN2 also_RR serve_VV0 for_IF verification_NN1 of_IO sense-making_NN1 in_II the_AT pLDA/K-means_FU clustering_NN1 processes_NN2 ._. 
</s>
<s>
In_II our_APPGE approach_NN1 ,_, the_AT extend_VV0 algorithm_NN1 does_VDZ not_XX only_RR rely_VVI on_II the_AT similarity_NN1 score_NN1 of_IO the_AT candidate_NN1 mappings_NN2 (_( i.e._REX ,_, weighted_JJ nodes_NN2 in_II the_AT conflict_NN1 graph_NN1 )_) ,_, but_CCB also_RR on_II how_RGQ strong_JJ each_DD1 mapping_NN1 is_VBZ in_II conflict_NN1 with_IW the_AT others_NN2 (_( i.e._REX ,_, degree_NN1 of_IO nodes_NN2 in_II the_AT graph_NN1 )_) ._. 
</s>
<s>
The_AT curve_NN1 is_VBZ shown_VVN in_II the_AT same_DA figure_NN1 as_CSA a_AT1 continuous_JJ line_NN1 and_CC is_VBZ normalized_VVN to_II 1_MC1 ._. 
</s>
<s>
It_PPH1 replies_VVZ to_II this_DD1 request_NN1 with_IW a_AT1 Status_NN1 200_MC OK_JJ response_NN1 indicating_VVG that_CST the_AT graph_NN1 concerning_II resource_NN1 http_NNU :_: had_VHD successfully_RR been_VBN updated_VVN ._. 
</s>
<s>
To_TO present_VVI our_APPGE results_NN2 ,_, we_PPIS2 used_VVD a_AT1 dumb-bell_JJ topology_NN1 as_CSA shown_VVN inFig._NNU 6to_FO simulate_VV0 a_AT1 bottleneck_NN1 link_NN1 shared_VVN by_II two_MC flows_NN2 ._. 
</s>
<s>
We_PPIS2 hope_VV0 that_CST this_DD1 study_NN1 will_VM inspire_VVI future_JJ developers_NN2 to_TO carefully_RR input_VVI effort_NN1 data_NN that_CST can_VM be_VBI used_VVN for_IF future_JJ studies_NN2 ._. 
</s>
<s>
Voinov_NP1 and_CC Cerco_NP1 (_( 2010_MC )_) and_CC Voinov_NP1 and_CC Shugart_NP1 (_( 2013_MC )_) discuss_VV0 these_DD2 issues_NN2 and_CC point_VV0 out_RP several_DA2 challenges_NN2 and_CC potential_JJ pitfalls_NN2 regarding_II the_AT construction_NN1 of_IO IEM_NN1 systems_NN2 ._. 
</s>
<s>
All_DB the_AT experiments_NN2 have_VH0 been_VBN performed_VVN using_VVG a_AT1 Pentium-IV_JJ PC_NN1 ,_, with_IW CPU_NN1 speed_NN1 of_IO 2.0_MC GHz_NNU ,_, 1_MC1 GB_NP1 RAM_NN1 and_CC MATLAB_VV0 7.0_MC ._. 
</s>
<s>
The_AT most_RGT important_JJ particular_JJ ,_, from_II the_AT perspective_NN1 of_IO the_AT IPAP-SA_NP1 ,_, is_VBZ the_AT person_NN1 to_TO be_VBI monitored_VVN ._. 
</s>
<s>
This_DD1 feature_NN1 provides_VVZ an_AT1 effective_JJ way_NN1 to_TO thwart_VVI denial-of-service_JJ attacks_NN2 from_II malicious_JJ CGI_NP1 scripts_NN2 ,_, or_CC poorly_RR written_VVN CGI_NP1 scripts_NN2 that_CST end_VV0 up_RP consuming_VVG system_NN1 resources_NN2 to_II the_AT detriment_NN1 of_IO the_AT other_JJ e-commerce_JJ services_NN2 ._. 
</s>
<s>
As_II21 for_II22 losses_NN2 ,_, the_AT maximum_JJ allowable_JJ loss_NN1 rate_NN1 is_VBZ 1.5%_FO for_IF an_AT1 acceptable_JJ VoIP_NN1 call_NN1 (_( Shevtekar_NP1 and_CC Ansari_NP1 ,_, 2006_MC )_) ._. 
</s>
<s>
Further_RRR ,_, different_JJ assessment_NN1 techniques_NN2 are_VBR probably_RR needed_VVN across_II the_AT six_MC areas_NN2 ._. 
</s>
<s>
The_AT variables_NN2 contained_VVN in_II dom(1)dom(2)_FO are_VBR called_VVN join_VV0 variables_NN2 ._. 
</s>
<s>
The_AT so-obtained_JJ Student_NN1 '_NULL s-t_NNU hidden_VVN Markov_NP1 model_NN1 (_( SHMM_NP1 )_) has_VHZ been_VBN considered_VVN in_II &lsqb;_( 4_MC &rsqb;_) under_II the_AT ML_NNU paradigm_NN1 using_VVG the_AT EM_FU algorithm_NN1 ;_; as_CSA it_PPH1 has_VHZ been_VBN shown_VVN ,_, the_AT SHMM_NP1 provides_VVZ an_AT1 effective_JJ ,_, computationally_RR efficient_JJ and_CC application-independent_JJ means_NN for_IF outlier_JJR tolerant_JJ representation_NN1 and_CC classification_NN1 of_IO sequential_JJ data_NN by_II31 means_II32 of_II33 continuous_JJ HMMs_NP1 ._. 
</s>
<s>
We_PPIS2 use_VV0 the_AT algorithm_NN1 described_VVN in_II Section_NN1 6.3_MC for_IF feature_NN1 selection_NN1 ._. 
</s>
<s>
According_II21 to_II22 our_APPGE discussion_NN1 with_IW managers_NN2 in_II the_AT PC_NN1 industry_NN1 and_CC other_JJ industries_NN2 ,_, these_DD2 professionals_NN2 are_VBR highly_RR involved_JJ in_II supplier_NN1 selection_NN1 ,_, and_CC often_RR are_VBR part_NN1 of_IO the_AT product_NN1 management_NN1 team_NN1 that_CST decides_VVZ whether_CSW to_TO manufacture_VVI a_AT1 product_NN1 in-house_JJ or_CC to_TO outsource_VVI ._. 
</s>
<s>
Overall_RR ,_, seven_MC of_IO the_AT 24_MC modified_JJ methods_NN2 are_VBR located_VVN in_II theIRFactoryclass_NN1 ,_, including_II the_AT three_MC ranked_JJ highest_JJT by_II our_APPGE approach_NN1 ._. 
</s>
<s>
Each_DD1 delta_NN1 chain_NN1 consists_VVZ of_IO two_MC dictionaries_NN2 ,_, one_MC1 for_IF the_AT snapshot_NN1 and_CC one_PN1 for_IF the_AT deltas_NN2 ._. 
</s>
<s>
In_II some_DD systems_NN2 ,_, the_AT bit-stream_NN1 can_VM be_VBI modified_VVN remotely_RR ,_, and_CC authentication_NN1 mechanisms_NN2 should_VM be_VBI employed_VVN to_TO prevent_VVI unauthorized_JJ users_NN2 from_II uploading_VVG a_AT1 malicious_JJ design_NN1 ,_, which_DDQ could_VM change_VVI the_AT intended_JJ functionality_NN1 of_IO the_AT device_NN1 ._. 
</s>
<s>
As_CSA discussed_VVN in_II Section_NN1 3.4_MC ,_, one_MC1 drawback_NN1 of_IO the_AT HMM-based_JJ method_NN1 is_VBZ that_CST HMMs_NP1 are_VBR trained_VVN generatively_RR using_VVG only_RR the_AT positive_JJ examples_NN2 of_IO the_AT symbol_NN1 ._. 
</s>
<s>
Only_RR by_II exploring_VVG the_AT implications_NN2 of_IO integrating_VVG global_JJ agricultural_JJ systems_NN2 ,_, energy_NN1 systems_NN2 and_CC carbon_NN1 price_NN1 schemes_NN2 ,_, can_VV0 a_AT1 comprehensive_JJ understanding_NN1 of_IO the_AT profound_JJ implications_NN2 of_IO climate_NN1 change_NN1 for_IF agriculture_NN1 and_CC global_JJ food_NN1 security_NN1 be_VBI achieved_VVN ._. 
</s>
<s>
First_MD ,_, although_CS we_PPIS2 used_VVD over_RG a_AT1 hundred_NNO interviews_NN2 with_IW individuals_NN2 who_PNQS occupied_VVD different_JJ organizational_JJ roles_NN2 ,_, in_II all_DB likelihood_NN1 we_PPIS2 did_VDD not_XX generate_VVI a_AT1 truly_RR exhaustive_JJ list_NN1 of_IO end_NN1 user_NN1 security_NN1 behaviors_NN2 ._. 
</s>
<s>
Table_NN1 8_MC and_CC Table_NN1 7_MC show_VV0 the_AT predictive_JJ and_CC explanative_JJ power_NN1 of_IO each_DD1 model_NN1 ._. 
</s>
<s>
We_PPIS2 have_VH0 added_VVN these_DD2 rules_NN2 to_II the_AT ASD_NN1 shown_VVN in_II Fig._NN1 13_MC to_TO show_VVI how_RRQ such_DA changes_NN2 will_VM affect_VVI activities_NN2 4_MC and_CC 5_MC of_IO our_APPGE methodology_NN1 (_( Fig._NN1 9_MC )_) ._. 
</s>
<s>
Most_DAT of_IO the_AT documentation_NN1 addresses_VVZ the_AT issue_NN1 of_IO availability_NN1 and_CC support_NN1 ._. 
</s>
<s>
When_CS analytical_JJ solutions_NN2 are_VBR not_XX available_JJ ,_, the_AT discrete_JJ solution_NN1 on_II the_AT finest_JJT mesh_NN1 in_II the_AT multilevel_NN1 hierarchy_NN1 is_VBZ used_VVN as_II the_AT reference_NN1 solution_NN1 and_CC coarser_JJR solutions_NN2 are_VBR projected_VVN onto_II the_AT finest_JJT mesh_NN1 in_BCL21 order_BCL22 to_TO approximate_VVI error_NN1 ._. 
</s>
<s>
Here_RL ,_, we_PPIS2 note_VV0 that_CST both_DB2 Cmax_VV0 and_CC C2_FO share_VV0 the_AT common_JJ term_NN1 of_IO Cp1_FO ._. 
</s>
<s>
The_AT Negotiator_NN1 is_VBZ a_AT1 customer_NN1 representative_NN1 that_CST has_VHZ in-depth_JJ domain_NN1 knowledge_NN1 ,_, provides_VVZ requirements_NN2 to_II the_AT development_NN1 team_NN1 ,_, and_CC is_VBZ willing_JJ to_TO carry_VVI responsibility_NN1 of_IO project_NN1 success_NN1 or_CC failure_NN1 ._. 
</s>
<s>
We_PPIS2 refer_VV0 to_II this_DD1 approach_NN1 asbacktracking_VVG ._. 
</s>
<s>
The_AT presence_NN1 of_IO the_AT soil_NN1 state_NN1 variable_NN1 is_VBZ decisive_JJ in_II configuring_VVG the_AT phase_NN1 space_NN1 of_IO the_AT model_NN1 ._. 
</s>
<s>
Nevertheless_RR ,_, we_PPIS2 believe_VV0 that_CST the_AT research_NN1 reported_VVN in_II this_DD1 study_NN1 is_VBZ an_AT1 important_JJ step_NN1 in_II the_AT empirical_JJ research_NN1 of_IO demographic_JJ targeting_VVG in_II the_AT sponsored_JJ search_NN1 area_NN1 ._. 
</s>
<s>
There_EX are_VBR a_AT1 number_NN1 of_IO ontologies_NN2 that_CST have_VH0 been_VBN specifically_RR developed_VVN to_TO support_VVI provenance_NN1 within_II Linked_JJ Data_NN ._. 
</s>
<s>
These_DD2 contours_NN2 ,_, called_VVN risk_NN1 contour_NN1 plots_NN2 in_II this_DD1 study_NN1 ,_, represent_VV0 the_AT probability_NN1 that_CST the_AT beach_NN1 area_NN1 is_VBZ less_DAR than_CSN a_AT1 certain_JJ value_NN1 in_II a_AT1 given_JJ year_NNT1 ._. 
</s>
<s>
In_II contrast_NN1 ,_, for_IF all_DB other_JJ DG_NP1 solutions_NN2 shown_VVN which_DDQ have_VH0 a_AT1 positive_JJ value_NN1 for_IF the_AT stabilization_NN1 parameter_NN1 ,_, as_RG well_RR as_CSA for_IF the_AT one_PN1 obtained_VVN with_IW a_AT1 piecewise_JJ linear_JJ ,_, conforming_VVG approximation_NN1 ,_, the_AT convergence_NN1 rate_NN1 is_VBZ approximately_RR h2h2_FO ._. 
</s>
<s>
Individual_JJ techniques_NN2 such_II21 as_II22 keystroke_NN1 analysis_NN1 can_VM provide_VVI valuable_JJ enhancements_NN2 in_II certain_JJ contexts_NN2 ,_, but_CCB are_VBR not_XX suited_VVN to_II all_DB users_NN2 and_CC scenarios_NN2 ._. 
</s>
<s>
When_CS queried_VVN with_IW a_AT1 seed_NN1 speaker_NN1 model_NN1 ,_, this_DD1 index_NN1 will_VM return_VVI a_AT1 reasonably_RR small_JJ list_NN1 of_IO candidate_NN1 speakers_NN2 that_CST may_VM correspond_VVI to_II the_AT same_DA person_NN1 ,_, which_DDQ can_VM be_VBI used_VVN to_TO greatly_RR reduce_VVI the_AT number_NN1 of_IO comparisons_NN2 needed_VVN to_TO cluster_VVI these_DD2 models_NN2 ._. 
</s>
<s>
The_AT images_NN2 come_VV0 from_II Walther_NP1 and_CC Koch_NP1 &lsqb;_( 5_MC &rsqb;_) 2_MC and_CC the_AT Berkeley_NP1 Segmentation_NN1 Dataset_NN1 &lsqb;_( 20_MC &rsqb;_) ._. 
</s>
<s>
In_II31 terms_II32 of_II33 the_AT pressures_NN2 for_IF journal_NN1 selectivity_NN1 there_EX is_VBZ mixed_VVN evidence_NN1 ._. 
</s>
<s>
Identify_VV0 the_AT dominating_JJ peaks_NN2 in_II Tr(i)_NP1 ,_, Tg(i)_NP1 and_CC Tb(i)_NN1 by_II examining_VVG the_AT turning_NN1 point_NN1 which_DDQ having_VHG positive_JJ to_II negative_JJ gradient_NN1 change_NN1 and_CC the_AT number_NN1 of_IO pixels_NN2 is_VBZ greater_JJR than_CSN a_AT1 predefined_JJ threshold_NN1 ,_, H._NP1 The_AT editors_NN2 '_NULL reported_VVD time_NNT1 to_TO reformulate_VVI the_AT particular_JJ question_NN1 thread_NN1 was_VBDZ 5_MC ?_? min_NNU ._. 
</s>
<s>
The_AT end_NN1 goal_NN1 of_IO SocioSpaces_NP2 is_VBZ to_TO enhance_VVI the_AT users_NN2 '_NULL experience_VV0 by_II delivering_VVG services_NN2 that_CST are_VBR meaningful_JJ and_CC useful_JJ to_II them_PPHO2 ._. 
</s>
<s>
The_AT proper_JJ way_NN1 to_TO deal_VVI with_IW this_DD1 issue_NN1 is_VBZ to_TO use_VVI high-order_JJ geometric_JJ approximation_NN1 of_IO the_AT boundary_NN1 ._. 
</s>
<s>
Using_VVG appearance_NN1 hashing_NN1 will_VM produce_VVI higher_JJR Recall_NN1 and_CC Precision_NN1 ._. 
</s>
<s>
This_DD1 deviates_VVZ from_II the_AT conventional_JJ approach_NN1 where_CS uncertainty_NN1 analysis_NN1 is_VBZ performed_VVN in_II the_AT inventory_NN1 phase_NN1 ._. 
</s>
<s>
The_AT perceived_JJ cues_NN2 are_VBR then_RT processed_VVN and_CC a_AT1 judgment_NN1 is_VBZ made_VVN about_II what_DDQ is_VBZ being_VBG viewed_VVN ._. 
</s>
<s>
In_II practice_NN1 ,_, the_AT designer_NN1 chooses_VVZ a_AT1 model_NN1 that_CST he_PPHS1 or_CC she_PPHS1 feels_VVZ will_VM adequately_RR represent_VVI the_AT behavior_NN1 of_IO the_AT built_VVN system_NN1 as_II31 well_II32 as_II33 its_APPGE future_JJ excitation_NN1 ;_; however_RR ,_, there_EX is_VBZ always_RR uncertainty_NN1 about_II which_DDQ values_NN2 of_IO the_AT model_NN1 parameters_NN2 will_VM give_VVI the_AT best_JJT representation_NN1 of_IO the_AT constructed_JJ system_NN1 and_CC its_APPGE environment_NN1 ,_, so_CS this_DD1 parameter_NN1 uncertainty_NN1 should_VM be_VBI quantified_VVN ._. 
</s>
<s>
In_BCL21 order_BCL22 to_TO benefit_VVI from_II this_DD1 definition_NN1 ,_, these_DD2 authors_NN2 developed_VVD a_AT1 four_MC dimensional_JJ framework_NN1 (_( 4-DAT_DD1 )_) to_TO crystallize_VVI the_AT key_JJ attributes_NN2 of_IO agility_NN1 :_: flexibility_NN1 ,_, speed_NN1 ,_, leanness_NN1 ,_, learning_VVG and_CC responsiveness_NN1 ._. 
</s>
<s>
Also_RR ,_, constraints_NN2 (_( 1_MC1 )_) and_CC (_( 2_MC )_) can_VM be_VBI combined_VVN into_II '_NULL '_NULL Sample_NN1 ._. 
</s>
<s>
We_PPIS2 investigate_VV0 the_AT stability_NN1 of_IO assignments_NN2 and_CC evaluate_VV0 the_AT costs_NN2 of_IO these_DD2 schemes_NN2 using_VVG real_JJ BGP_NP1 trace_NN1 data_NN in_II Section5_FO ._. 
</s>
<s>
Note_VV0 that_DD1 radio_NN1 basis_NN1 function_NN1 (_( RBF_NP1 )_) was_VBDZ adopted_VVN as_II the_AT kernel_NN1 function_NN1 for_IF implementing_VVG KDDA_NN1 ._. 
</s>
<s>
The_AT option_NN1 for_IF a_AT1 virtualization_NN1 solution_NN1 offers_VVZ the_AT greatest_JJT level_NN1 of_IO control_NN1 over_II organizational_JJ data_NN that_CST are_VBR used_VVN by_II employees_NN2 from_II their_APPGE devices_NN2 ._. 
</s>
<s>
The_AT lack_NN1 of_IO involvement_NN1 from_II the_AT business_NN1 and_CC residential_JJ communities_NN2 may_VM have_VHI biased_VVN the_AT study_NN1 in_II at_RR21 least_RR22 two_MC ways_NN2 ._. 
</s>
<s>
However_RR ,_, the_AT question_NN1 is_VBZ how_RRQ to_TO effectively_RR apply_VVI and_CC integrate_VVI high-level_JJ ArchiMate_NN1 with_IW other_JJ low-level_JJ modelling_NN1 standards_NN2 for_IF supporting_VVG end-to-end_JJ agile_JJ EA_NP1 modelling_VVG both_RR at_II the_AT high_JJ and_CC low_JJ detailed_JJ level_NN1 ._. 
</s>
<s>
There_EX are_VBR basic_JJ calls_NN2 that_CST must_VM happen_VVI in_II all_DB examples_NN2 of_IO microphone_NN1 recording_NN1 programs_NN2 that_CST determine_VV0 where_RRQ to_TO set_VVI the_AT hooks_NN2 ._. 
</s>
<s>
Knowledge_NN1 created_VVN is_VBZ subsequently_RR managed_VVN post-implementation_NN1 ,_, transferred_VVN ,_, then_RT retained_VVN by_II the_AT organization_NN1 ,_, and_CC ultimately_RR applied_VVN throughout_II the_AT ES_NP1 lifecycle_NN1 ._. 
</s>
<s>
After_II eliminating_VVG all_DB of_IO them_PPHO2 ,_, we_PPIS2 obtain_VV0 a_AT1 set_NN1 of_IO so_JJ21 called_JJ22 irreducible_JJ invariants_NN2 ._. 
</s>
<s>
The_AT histogram_NN1 in_II Fig._NN1 2_MC illustrates_VVZ the_AT observations_NN2 of_IO specific_JJ devices_NN2 for_IF a_AT1 user_NN1 throughout_II the_AT duration_NN1 of_IO their_APPGE 14-day_JJ experimental_JJ participation_NN1 ._. 
</s>
<s>
Driving_VVG up_II21 to_II22 an_AT1 access_NN1 point_NN1 each_DD1 time_NNT1 is_VBZ too_RG time_NNT1 consuming_VVG and_CC traffic_NN1 congestion_NN1 prone_JJ ._. 
</s>
<s>
A_AT1 major_JJ challenge_NN1 in_II the_AT area_NN1 of_IO attribution_NN1 is_VBZ that_DD1 of_IO prepositioning_VVG ._. 
</s>
<s>
The_AT idea_NN1 is_VBZ here_RL that_CST an_AT1 adversary_NN1 can_VM read_VVI messages_NN2 sent_VVN over_II the_AT network_NN1 and_CC collect_VV0 them_PPHO2 in_II her_APPGE knowledge_NN1 set_NN1 ._. 
</s>
<s>
The_AT curve_NN1 shows_VVZ the_AT trade-off_NN1 between_II true_JJ positives_NN2 and_CC false_JJ positives_NN2 as_II the_AT classification_NN1 threshold_NN1 parameter_NN1 within_II the_AT filter_NN1 is_VBZ varied_VVN ._. 
</s>
<s>
The_AT aggregated_JJ socio-economic_JJ output_NN1 indicators_NN2 could_VM be_VBI coupled_VVN with_IW relevant_JJ environmental_JJ consequences_NN2 ._. 
</s>
<s>
Table_NN1 8_MC shows_VVZ the_AT test_NN1 results_NN2 over_II the_AT AR_UH database_NN1 for_IF the_AT LMG_NP1 method_NN1 and_CC methods_NN2 in_II &lsqb;_( 32_MC &rsqb;_) ._. 
</s>
<s>
In_BCL21 order_BCL22 to_TO integrate_VVI these_DD2 two_MC datasets_NN2 ,_, diseases_NN2 and_CC genes_NN2 (_( including_II their_APPGE synonyms_NN2 )_) found_VVD in_II Linked_JJ TCGA_NN1 are_VBR required_VVN to_TO be_VBI identified_VVN in_II PubMed_JJ articles_NN2 metadata_NN1 ._. 
</s>
<s>
In_II this_DD1 paper_NN1 the_AT range_NN1 of_IO applicability_NN1 of_IO the_AT smoothing_JJ technique_NN1 is_VBZ extended_VVN by_II applying_VVG it_PPH1 to_II a_AT1 kinematic_JJ limit_NN1 analysis_NN1 formulation_NN1 incorporating_VVG error_NN1 estimation_NN1 ._. 
</s>
<s>
Regarding_II the_AT association_NN1 of_IO the_AT features_NN2 a_AT1 similar_JJ process_NN1 is_VBZ performed_VVN for_IF all_DB the_AT features_NN2 belonging_VVG to_II the_AT same_DA planar_JJ region_NN1 ._. 
</s>
<s>
The_AT OWL-S_NN2 document_VV0 maps_NN2 each_DD1 operation_NN1 and_CC message_NN1 defined_VVN in_II the_AT WSDL_NP1 definition_NN1 to_II an_AT1 ontology_NN1 ._. 
</s>
<s>
As_II an_AT1 NSF-sponsored_JJ effort_NN1 ,_, GENI_NP1 is_VBZ widely_RR available_JJ to_II researchers_NN2 both_RR inside_RL and_CC outside_II the_AT U.S._NP1 It_PPH1 yields_VVZ stress_NN1 test_NN1 requirements_NN2 that_CST are_VBR constructed_VVN from_II specific_JJ control_NN1 flow_NN1 paths_NN2 along_II21 with_II22 time_NNT1 values_NN2 indicating_VVG when_RRQ to_TO trigger_VVI them_PPHO2 ._. 
</s>
<s>
Finally_RR ,_, several_DA2 multi-attribute_JJ models_NN2 of_IO quality_NN1 for_IF software_NN1 engineering_NN1 processes_NN2 are_VBR critically_RR reviewed_VVN to_TO establish_VVI the_AT justification_NN1 for_IF the_AT research_NN1 described_VVD herein_RR ._. 
</s>
<s>
These_DD2 factors_NN2 make_VV0 the_AT crawler_NN1 a_AT1 complex_JJ component_NN1 of_IO the_AT system_NN1 ._. 
</s>
<s>
In_II the_AT absence_NN1 of_IO appropriate_JJ theoretical_JJ background_NN1 ,_, findings_NN2 from_II a_AT1 LCCS_NP2 can_VM be_VBI used_VVN to_TO generate_VVI theoretical_JJ and_CC conceptual_JJ structures_NN2 ,_, such_II21 as_II22 conjectures_NN2 ,_, propositions_NN2 ,_, hypotheses_NN2 ,_, and_CC definitions_NN2 of_IO constructs_NN2 ._. 
</s>
<s>
All_DB the_AT scholar_NN1 representations_NN2 depend_VV0 on_II the_AT paper_NN1 and_CC the_AT citation_NN1 relationships_NN2 among_II them_PPHO2 ._. 
</s>
<s>
The_AT digitisation_NN1 project_NN1 was_VBDZ a_AT1 great_JJ success_NN1 but_CCB the_AT metadata_NN1 for_IF it_PPH1 was_VBDZ of_IO limited_JJ quality_NN1 and_CC quantity_NN1 ._. 
</s>
<s>
For_REX21 example_REX22 ,_, if_CS the_AT first_MD of_IO several_DA2 fraudulent_JJ transactions_NN2 occurs_VVZ at_II 10am_RA and_CC signature_NN1 processing_NN1 is_VBZ implemented_VVN within_II a_AT1 daily_JJ processing_NN1 model_NN1 ,_, these_DD2 instances_NN2 only_RR be_VBI identified_VVN at_II the_AT end_NN1 of_IO the_AT current_JJ financial_JJ trading_NN1 day_NNT1 ._. 
</s>
<s>
Therefore_RR ,_, the_AT virtual_JJ lock_NN1 (_( a_AT1 physical_JJ lock_NN1 does_VDZ not_XX actually_RR exist_VVI )_) is_VBZ acquired_VVN after_II the_AT user_NN1 finishes_VVZ his/her_APPGE work_NN1 ._. 
</s>
<s>
Furthermore_RR ,_, a_AT1 cell-to-cell_JJ map_NN1 comparison_NN1 is_VBZ even_RR less_RGR effective_JJ when_CS analysing_VVG patterns_NN2 and_CC shapes_NN2 of_IO urban_JJ growth_NN1 ,_, where_CS it_PPH1 is_VBZ more_RGR appropriate_JJ to_TO study_VVI how_RRQ shape_NN1 of_IO urban_JJ fragments_NN2 (_( patches_NN2 formed_VVN by_II contiguous_JJ cells_NN2 showing_VVG the_AT same_DA use_NN1 )_) varies_VVZ between_II one_MC1 map_NN1 and_CC another_DD1 (_( Li_NP1 et_NP1 ?_? al._NNU ,_, 2008_MC )_) ._. 
</s>
<s>
The_AT blocking_JJ probability_NN1 is_VBZ pn+k_FO where_RRQ requests_NN2 that_CST can_VM not_XX be_VBI deferred_VVN (_( arriving_VVG after_II the_AT system_NN1 is_VBZ in_II state_NN1 n+k_FO )_) are_VBR rejected_VVN ._. 
</s>
<s>
Each_DD1 snapshot_NN1 triple_NN1 is_VBZ queried_VVN within_II the_AT deletion_NN1 tree_NN1 ._. 
</s>
<s>
ArchiMate_NN1 is_VBZ used_VVN to_TO demonstrate_VVI the_AT modelling_NN1 of_IO infrastructure_NN1 architecture_NN1 that_CST is_VBZ required_VVN to_TO support_VVI the_AT procurement_NN1 management_NN1 application_NN1 such_II21 as_II22 order_NN1 processing_NN1 ,_, financial_JJ and_CC product_NN1 management_NN1 applications_NN2 ._. 
</s>
<s>
This_DD1 roving_JJ bugnet_NN1 design_NN1 is_VBZ presented_VVN in_II two_MC layers_NN2 :_: the_AT OS_NP2 specific_JJ prototype_NN1 microphone_NN1 surveillance_NN1 program_NN1 in_II Section2.1_FO ;_; and_CC ,_, the_AT remote_JJ management_NN1 layer_NN1 which_DDQ is_VBZ comprised_VVN of_IO the_AT IRC_NN1 bot_NN1 and_CC botnet_NN1 in_II Section2.2_FO ._. 
</s>
<s>
RSS_NP1 is_VBZ an_AT1 XML-based_JJ specification_NN1 format_VV0 that_DD1 is_VBZ mainly_RR used_JJ to_II syndicate_NN1 news_NN1 sites_NN2 ,_, personal_JJ Weblogs_NN2 ,_, CVS_NN2 checkins_VVZ ,_, etc_RA ._. 
</s>
<s>
However_RR ,_, this_DD1 procedure_NN1 needs_VVZ an_AT1 efficient_JJ and_CC effective_JJ search_NN1 algorithm_NN1 to_TO find_VVI this_DD1 subset_NN1 of_IO parameter_NN1 combinations_NN2 ._. 
</s>
<s>
Tangent_JJ sigmoid_JJ functions_NN2 have_VH0 been_VBN used_VVN in_II the_AT hidden_JJ layers_NN2 as_CSA transfer_NN1 functions_NN2 ,_, while_CS the_AT output_NN1 layer_NN1 uses_VVZ a_AT1 linear_JJ function_NN1 ._. 
</s>
<s>
Rule_NN1 axioms_NN2 are_VBR similar_JJ to_II OWL_NN1 axioms_NN2 ,_, except_CS they_PPHS2 have_VH0 as_II their_APPGE element_NN1 name_NN1 ._. 
</s>
<s>
The_AT queries_NN2 are_VBR selected_VVN in_II such_DA a_AT1 way_NN1 that_CST they_PPHS2 will_VM be_VBI evaluated_VVN over_II triples_NN2 of_IO a_AT1 certain_JJ dynamicity_NN1 ,_, which_DDQ requires_VVZ the_AT benchmarked_JJ systems_NN2 to_TO handle_VVI this_DD1 dynamicity_NN1 well_RR ._. 
</s>
<s>
To_TO maximize_VVI the_AT likely_JJ effectiveness_NN1 of_IO outcomes_NN2 ,_, we_PPIS2 used_VVD a_AT1 set_NN1 of_IO interviews_NN2 to_TO elicit_VVI practitioners_NN2 '_NULL opinions_NN2 about_II behaviors_NN2 of_IO concern_NN1 ,_, so_CS21 that_CS22 we_PPIS2 could_VM focus_VVI on_II those_DD2 perceived_VVN as_CSA most_DAT significant_JJ ._. 
</s>
<s>
Taken_VVN together_RL ,_, associating_VVG our_APPGE contextual_JJ analysis_NN1 over_II time_NNT1 with_IW our_APPGE task_NN1 change_NN1 analysis_NN1 confirms_VVZ that_DD1 core_NN1 developers_NN2 ,_, although_CS less_RGR active_JJ in_II the_AT final_JJ project_NN1 phase_NN1 ,_, were_VBDR indeed_RR above_II average_JJ performers_NN2 ,_, and_CC these_DD2 members_NN2 were_VBDR appropriately_RR selected_VVN &lsqb;_( 10_MC &rsqb;_) ._. 
</s>
<s>
In_II Section2_FO ,_, we_PPIS2 define_VV0 agility_NN1 ,_, based_VVN on_II earlier_JJR studies&lsqb;4&rsqb;_FO ,_, then_RT summarize_VV0 the_AT 4-DAT_DD1 (_( based_VVN on&lsqb;3&rsqb;_FO )_) ._. 
</s>
<s>
This_DD1 may_VM be_VBI undesirable_JJ from_II the_AT perspective_NN1 of_IO Client1_FO ._. 
</s>
<s>
The_AT experiment_NN1 steps_NN2 ,_, training_NN1 and_CC output_NN1 produced_VVN during_II the_AT experiment_NN1 steps_NN2 are_VBR provided_VVN in_II Table_NN1 2_MC ._. 
</s>
<s>
The_AT corresponding_JJ Ekman_NN1 spiral_NN1 to_II Example_NN1 2._MC b_ZZ1 is_VBZ depicted_VVN in_II Fig._NN1 5_MC ._. 
</s>
<s>
They_PPHS2 all_DB are_VBR memory-based_JJ hash_NN1 tables_NN2 with_IW varying_JJ values_NN2 attached_VVN to_II each_DD1 word_NN1 ._. 
</s>
<s>
More_RGR recently_RR ,_, HQ_NN1 dataset_NN1 was_VBDZ developed_VVN for_IF surface_NN1 humidity_NN1 (_( dew-point_JJ daily_JJ maximum_NN1 and_CC minimum_NN1 :_: dTmin_NN1 and_CC dTmax_NN1 )_) by_II Lucas_NP1 (_( 2006_MC )_) for_IF the_AT period_NN1 1957–2003_MCMC and_CC for_IF surface_NN1 pan_NN1 evaporation_NN1 (_( pE_NN1 )_) (_( Jovanovic_NN1 et@al._FO ,_, 2008_MC )_) for_IF the_AT period_NN1 1975–2003_MCMC ._. 
</s>
<s>
We_PPIS2 train_VV0 a_AT1 model_NN1 on_II tunnel_NN1 traffic_NN1 generated_VVN by_II a_AT1 broad_JJ range_NN1 of_IO HTTP_NP1 tunnelling_NN1 applications_NN2 and_CC also_RR on_II representative_JJ background_NN1 normal_JJ traffic_NN1 ._. 
</s>
<s>
Predicting_VVG fire_NN1 effects_NN2 is_VBZ subject_II21 to_II22 multiple_JJ sources_NN2 of_IO uncertainty_NN1 ,_, including_II limited_JJ or_CC inadequate_JJ empirical_JJ observations_NN2 ,_, and_CC gaps_NN2 in_II fire_NN1 effects_NN2 science_NN1 (_( Hyde_NP1 et_RA21 al._RA22 ,_, 2012_MC )_) ,_, highlighting_VVG a_AT1 need_NN1 for_IF continued_JJ empirical_JJ research_NN1 into_II post-fire_JJ impacts_NN2 (_( Riley_NP1 et_RA21 al._RA22 ,_, 2013_MC )_) ._. 
</s>
<s>
The_AT resulting_JJ algorithm_NN1 offers_VVZ linear_JJ computational_JJ complexity_NN1 in_II the_AT number_NN1 of_IO spacetime_NNT1 elements_NN2 ._. 
</s>
<s>
Confidence_NN1 refers_VVZ to_II the_AT lowest_JJT threshold_NN1 of_IO certainty_NN1 which_DDQ the_AT system_NN1 must_VM have_VHI in_BCL21 order_BCL22 to_TO return_VVI an_AT1 annotation_NN1 ._. 
</s>
<s>
The_AT negative_JJ effects_NN2 of_IO such_DA guidance_NN1 can_VM not_XX be_VBI predicted_VVN in_II advance_NN1 of_IO experiment_NN1 and_CC would_VM not_XX be_VBI discovered_VVN until_CS sufficient_JJ data_NN are_VBR explored_VVN ._. 
</s>
<s>
Considering_CS these_DD2 two_MC adages_NN2 together_RL in_II the_AT context_NN1 of_IO e-commerce_JJ security_NN1 ,_, privacy_NN1 and_CC client_NN1 trust_NN1 ,_, it_PPH1 becomes_VVZ clear_JJ that_CST malicious_JJ perpetrators_NN2 will_VM rarely_RR attempt_VVI to_TO break_VVI encryption_NN1 codes_NN2 when_CS they_PPHS2 can_VM much_RR more_RGR easily_RR break_VVI into_II a_AT1 system_NN1 ,_, and_CC enjoy_VVI a_AT1 much_RR higher_JJR return_NN1 via_II this_DD1 simpler_JJR approach_NN1 (_( Ghosh_NP1 ,_, 2000_MC )_) ._. 
</s>
<s>
When_CS δ_FO is_VBZ set_VVN to_TO be_VBI smaller_JJR ,_, a_AT1 more_RGR uniform_JJ refinement_NN1 is_VBZ favoured_VVN ,_, which_DDQ affects_VVZ efficiency_NN1 ._. 
</s>
<s>
This_DD1 may_VM be_VBI true_JJ but_CCB it_PPH1 is_VBZ still_RR hard_JJ to_TO imagine_VVI how_RRQ a_AT1 painting_NN1 like_II Mona_NP1 Lisa_NP1 could_VM be_VBI created_VVN by_II Leonardo_NP1 working_VVG together_RL with_IW Michelangelo_NP1 and_CC Raphael_NP1 in_II a_AT1 community_NN1 effort_NN1 ._. 
</s>
<s>
The_AT additional_JJ requirement_NN1 is_VBZ that_CST transitionthas_NN2 no_AT reset_JJ arcs_NN2 ._. 
</s>
<s>
The_AT Eulerian_JJ framework_NN1 is_VBZ ,_, however_RR ,_, not_XX best_RRT suited_VVN for_IF riverine_JJ processes_NN2 that_CST evolve_VV0 both_RR in_II space_NN1 and_CC time_NNT1 (_( e.g._REX ,_, bedforms_NN2 ,_, coherent_JJ structures_NN2 ,_, plumes_NN2 )_) ,_, and_CC thus_RR observing_VVG from_II one_MC1 location_NN1 might_VM hinder_VVI capturing_VVG the_AT dynamics_NN of_IO the_AT process_NN1 as_CSA it_PPH1 unfolds_VVZ ._. 
</s>
<s>
Hence_RR ,_, parallel_JJ communication_NN1 functions_NN2 near_II the_AT computation_NN1 loop_NN1 are_VBR prime_JJ candidates_NN2 for_IF adaptation_NN1 control_NN1 points_VVZ in_II using_VVG our_APPGE methods_NN2 ._. 
</s>
<s>
Results_NN2 for_IF HMMPayl_NP1 (_( Ariu_NP1 et_RA21 al._RA22 ,_, 2011_MC )_) for_IF the_AT CLET_NN1 dataset_NN1 are_VBR given_VVN in_II Table_NN1 11_MC ,_, where_CS the_AT average_JJ AUCp_NN1 value_NN1 is_VBZ given_VVN for_IF each_DD1 rule_NN1 used_VMK to_TO combine_VVI multiple_NN1 HMM_UH scores_NN2 ._. 
</s>
<s>
Finally_RR ,_, relevant_JJ software_NN1 project_NN1 success_NN1 indicators_NN2 include_VV0 measures_NN2 related_VVN to_II software_NN1 impact_NN1 on_II the_AT development_NN1 organization_NN1 ,_, the_AT reviews_NN2 of_IO post-release_JJ customers_NN2 ,_, and_CC actual_JJ software_NN1 usage_NN1 &lsqb;_( 112_MC &rsqb;_) ._. 
</s>
<s>
The_AT proposed_JJ approach_NN1 for_IF the_AT calculation_NN1 of_IO probabilities_NN2 of_IO failure_NN1 is_VBZ applied_VVN to_II two_MC problems_NN2 ._. 
</s>
<s>
The_AT nature_NN1 and_CC relatively_RR small_JJ size_NN1 of_IO the_AT sample_NN1 limit_VV0 the_AT capacity_NN1 to_TO generalize_VVI research_NN1 findings_NN2 across_II all_DB types_NN2 of_IO business_NN1 organizations_NN2 ._. 
</s>
<s>
A_AT1 balanced_JJ network_NN1 means_VVZ that_CST |_NULL Zi_NP1 |_NULL and_CC L_ZZ1 (_( Zi_NP1 ,_, Zi_NP1 )_) are_VBR kept_VVN the_AT same_DA for_IF all_DB i_ZZ1 ,_, and_CC L_ZZ1 (_( Zi_NP1 ,_, Zj_NP1 )_) is_VBZ kept_VVN the_AT same_DA for_IF all_DB pairs_NN2 of_IO clusters_NN2 i_ZZ1 and_CC j_ZZ1 ._. 
</s>
<s>
We_PPIS2 then_RT generated_VVD a_AT1 balanced_JJ random_JJ network_NN1 by_II randomly_RR drawing_VVG edges_NN2 between_II nodes_NN2 ,_, keeping_VVG these_DD2 three_MC constraints_NN2 the_AT same_DA ._. 
</s>
<s>
Modifiers_NN2 are_VBR specific_JJ types_NN2 of_IO patterns_NN2 that_CST can_VM extend_VVI ,_, refine_VVI or_CC determine_VVI a_AT1 variation_NN1 of_IO other_JJ software_NN1 patterns_NN2 (_( see_VV0 Kolfschoten_NP1 et_RA21 al._RA22 ,_, 2011_MC )_) when_CS used_VVN in_II combination_NN1 ._. 
</s>
<s>
Assuming_VVG the_AT frequency_NN1 has_VHZ deviated_VVN from_II the_AT nominal_JJ value_NN1 of_IO 60_MC Hz_NNU ,_, the_AT CCA_NP1 transmits_VVZ load_NN1 shed_NN1 measurement_NN1 to_II the_AT load_NN1 agent_NN1 (_( LA_FW )_) at_II ksuHost3_FO ._. 
</s>
<s>
The_AT density_NN1 difference_NN1 between_II layers_NN2 was_VBDZ 19.96@kg/m3_FU ._. 
</s>
<s>
Following_VVG scale_NN1 refinement_NN1 and_CC reliability_NN1 and_CC validity_NN1 testing_NN1 ,_, we_PPIS2 proceeded_VVD to_TO test_VVI the_AT fit_NN1 of_IO the_AT revised_JJ model_NN1 containing_VVG 29_MC items_NN2 ._. 
</s>
<s>
There_EX is_VBZ also_RR the_AT possibility_NN1 that_CST ,_, when_CS tested_VVN alongside_II general_JJ measures_NN2 of_IO teamwork_NN1 ,_, the_AT importance_NN1 for_IF performance_NN1 of_IO agile_JJ methods_NN2 is_VBZ found_VVN to_TO be_VBI subordinate_JJ to_II its_APPGE associated_JJ teamworking_NN1 ._. 
</s>
<s>
Contrarily_RR ,_, the_AT metadata_NN1 within_II MP4_FO or_CC MPEG-4_MC files_NN2 are_VBR usually_RR related_VVN to_II multiple_JJ camera_NN1 angles_NN2 ,_, or_CC different_JJ language_NN1 tracks_NN2 and_CC do_VD0 not_XX capture_VVI socioeconomic_JJ aspects_NN2 ._. 
</s>
<s>
The_AT highest_JJT layer_NN1 in_II the_AT hierarchy_NN1 tree_NN1 of_IO EMERALD_NN1 is_VBZ the_AT enterprise-wide_JJ layer_NN1 ,_, which_DDQ analyzes_VVZ misuse_NN1 across_II multiple_JJ domains_NN2 in_II the_AT entire_JJ system_NN1 ._. 
</s>
<s>
AlthoughTheorem_VV0 4does_FO not_XX requireAssumption_NN1 5_MC ,_, if_CS we_PPIS2 solveP_VV0 (_( H_ZZ1 ,_, N_ZZ1 ,_, K_ZZ1 )_) to_TO obtain_VVI a_AT1 design_NN1 minimizing_VVG the_AT average_JJ hierarchical_JJ routing_NN1 error_NN1 ,_, we_PPIS2 must_VM ensure_VVI that_CST the_AT choice_NN1 ofδ_FO (_( ·_FO )_) satisfiesAssumption_NN1 5_MC ._. 
</s>
<s>
This_DD1 may_VM cause_VVI individual_JJ ants_NN2 to_TO follow_VVI a_AT1 path_NN1 with_IW a_AT1 lower_JJR or_CC even_RR zero_MC pheromone_NN1 concentration_NN1 level_NN1 ._. 
</s>
<s>
Thus_RR ,_, the_AT voting_NN1 matrix_NN1 entry_NN1 for_IF a_AT1 pair_NN of_IO nodes_NN2 (_( A_ZZ1 ,_, B_ZZ1 )_) is_VBZ one_PN1 ._. 
</s>
<s>
The_AT remainder_NN1 of_IO the_AT paper_NN1 is_VBZ organized_VVN as_CSA follows_VVZ ._. 
</s>
<s>
Finally_RR ,_, in_II Section_NN1 5.7_MC ,_, we_PPIS2 mined_VVD the_AT pattern_NN1 dct:subject_FO some_DD skos:Concept_FO ._. 
</s>
<s>
A_AT1 one-tailed_JJ Wilcoxon_NP1 Signed_VVD Ranks_NN2 test_NN1 of_IO the_AT absolute_JJ residual_JJ rejects_NN2 both_RR null_JJ hypotheses_NN2 (_( p.035_FO and_CC p.0001_FO )_) so_CS one_PN1 can_VM be_VBI confident_JJ that_CST PEBA_NP1 EBA+_FO EBA++_FO is_VBZ not_XX a_AT1 chance_NN1 outcome_NN1 ._. 
</s>
<s>
As_II the_AT complexity_NN1 of_IO the_AT system_NN1 increases_NN2 ,_, the_AT performance_NN1 gap_NN1 between_II the_AT genetic_JJ algorithm_NN1 and_CC the_AT quasi-Newton_NP1 iteration_NN1 narrow_JJ and_CC ,_, for_IF a_AT1 large_JJ number_NN1 of_IO input_NN1 parameters_NN2 the_AT genetic_JJ algorithm_NN1 requires_VVZ fewer_DAR iterations_NN2 to_II convergence_NN1 ._. 
</s>
<s>
There_EX is_VBZ no_AT guarantee_NN1 about_II the_AT maximum_JJ size_NN1 of_IO blocks_NN2 ._. 
</s>
<s>
Up_II21 to_II22 this_DD1 point_NN1 ,_, I_PPIS1 have_VH0 discussed_VVN only_RR static_JJ structural_JJ properties_NN2 of_IO networkse.g._NNU ,_, degree_NN1 distribution_NN1 ,_, hubs_NN2 ,_, An_AT1 extension_NN1 of_IO this_DD1 detection-avoidance_JJ technique_NN1 would_VM be_VBI for_IF a_AT1 player_NN1 occasionally_RR to_TO choose_VVI sub-optimal_JJ moves_NN2 that_CST are_VBR still_RR a_AT1 good_JJ second_NNT1 or_CC third_MD choice_NN1 of_IO the_AT engine_NN1 ,_, leading_VVG to_II MM/CV_FU values_VVZ below_RG 100%_NNU ._. 
</s>
<s>
A_AT1 linear_JJ combination_NN1 of_IO native_JJ dimensions_NN2 within_II the_AT same_DA natural_JJ domain_NN1 can_VM be_VBI interpreted_VVN as_CSA an_AT1 index_NN1 of_IO the_AT information_NN1 contained_VVN in_II that_DD1 domain_NN1 ,_, which_DDQ reflects_VVZ the_AT impact_NN1 of_IO the_AT domain_NN1 on_II the_AT vulnerability_NN1 of_IO the_AT policy_NN1 under_II consideration_NN1 ._. 
</s>
<s>
Intangible_JJ benefits_NN2 are_VBR more_RGR relation-based_JJ and_CC may_VM be_VBI imparted_VVN through_II strong_JJ interpersonal_JJ bonds_NN2 with_IW the_AT provider_NN1 resulting_VVG in_II feelings_NN2 of_IO comfort_NN1 and_CC trust_NN1 ._. 
</s>
<s>
In_II all_DB these_DD2 runs_NN2 M=60_FO elements_NN2 were_VBDR used_VVN ._. 
</s>
<s>
For_REX21 example_REX22 ,_, the_AT Intercomm&lsqb;25&rsqb;_FO and_CC model_NN1 coupling_NN1 toolkit_NN1 &lsqb;_( 26_MC &rsqb;_) projects_NN2 provide_VV0 mechanisms_NN2 for_IF communication_NN1 and_CC interpolation_NN1 of_IO data_NN between_II two_MC grid-based_JJ applications_NN2 ._. 
</s>
<s>
For_REX21 example_REX22 ,_, an_AT1 offline_JJ study_NN1 may_VM suggest_VVI that_CST a_AT1 regression_NN1 model_NN1 with_IW degree_NN1 one_MC1 ,_, i.e._REX ,_, a1k+b1_FO ,_, should_VM be_VBI used_VVN ._. 
</s>
<s>
Furthermore_RR ,_, Google_NP1 is_VBZ a_AT1 complete_JJ architecture_NN1 for_IF gathering_VVG web_NN1 pages_NN2 ,_, indexing_VVG them_PPHO2 ,_, and_CC performing_VVG search_NN1 queries_NN2 over_II them_PPHO2 ._. 
</s>
<s>
Since_CS this_DD1 may_VM contain_VVI only_RR a_AT1 small_JJ fraction_NN1 of_IO all_DB the_AT variables_NN2 in_II the_AT original_JJ constraint_NN1 ,_, we_PPIS2 can_VM greatly_RR reduce_VVI the_AT amount_NN1 of_IO work_NN1 that_CST this_DD1 incremental_JJ propagation_NN1 requires_VVZ ._. 
</s>
<s>
The_AT MIS_NP2 evaluation_NN1 factor_NN1 represents_VVZ the_AT service_NN1 level_NN1 of_IO the_AT IS_VBZ and_CC security_NN1 groups_NN2 ._. 
</s>
<s>
It_PPH1 has_VHZ facilities_NN2 for_IF manual_JJ annotation_NN1 of_IO web_NN1 pages_NN2 but_CCB does_VDZ not_XX contain_VVI any_DD features_NN2 to_TO support_VVI automatic_JJ annotation_NN1 ._. 
</s>
<s>
According_II21 to_II22 Theorem_NN1 1_MC1 in_II Section_NN1 6.1.12_MC ,_, to_TO construct_VVI a_AT1 model_NN1 ,_, we_PPIS2 should_VM select_VVI one_MC1 and_CC only_JJ one_MC1 constraint_NN1 from_II each_DD1 subset_NN1 Si_FW in_II the_AT partitioned_JJ DS_NP1 ._. 
</s>
<s>
Other_JJ point_NN1 types_NN2 are_VBR rarely_RR accessed_VVN ,_, e.g._REX disconnecting_VVG the_AT home_NN1 or_CC generation_NN1 from_II the_AT grid_NN1 in_II this_DD1 article_NN1 '_NULL s_ZZ1 DSM_NP1 example_NN1 ._. 
</s>
<s>
Martins_NP2 and_CC Eloff_NP1 (_( 2002_MC )_) take_VV0 in_II consideration_NN1 the_AT user_NN1 '_NULL s_ZZ1 role_NN1 when_CS presenting_VVG a_AT1 model_NN1 for_IF implementing_VVG and_CC enhancing_VVG the_AT culture_NN1 of_IO IS_VBZ security_NN1 ._. 
</s>
<s>
As_CSA has_VHZ been_VBN discussed_VVN at_RR21 length_RR22 in_II the_AT literature_NN1 this_DD1 not_XX a_AT1 reasonable_JJ assumption_NN1 (_( e.g._REX ,_, IPCC_NP1 ,_, 2014_MC )_) ;_; 2_MC )_) impact_NN1 functions_NN2 such_II21 as_II22 Equation_NN1 (_( 14_MC )_) are_VBR calibrated_VVN using_VVG equilibrium_NN1 temperatures_NN2 ._. 
</s>
<s>
This_DD1 score_NN1 can_VM provide_VVI a_AT1 quick_JJ assessment_NN1 of_IO the_AT prospect_NN1 that_CST a_AT1 Japanese_JJ client_NN1 will_VM actually_RR decide_VVI to_TO outsource_VVI a_AT1 specific_JJ IT_NN1 project_NN1 ._. 
</s>
<s>
The_AT non-English_JJ participants_NN2 included_VVD :_: Mauritian_JJ (_( French_NP1 )_) ,_, Romanian_JJ ,_, Turkish_JJ ,_, Indian_JJ (_( Hindi_NP1 )_) ,_, Greek_NN1 ,_, Arabic_NN1 ,_, Burmese_NN ,_, Malay_JJ and_CC Indonesian_JJ ._. 
</s>
<s>
The_AT NP_NP1 values_NN2 show_VV0 the_AT extent_NN1 to_II which_DDQ the_AT land_NN1 use_NN1 analysed_VVN is_VBZ fragmented_JJ or_CC aggregated_JJ ._. 
</s>
<s>
Abstract_JJ |_NULL PDF_NP1 (_( 502_MC K_ZZ1 )_) |_NULL View_NN1 Record_NN1 in_II Scopus_NP1 |_NULL Cited_VVD By_II in_II Scopus_NP1 (_( 109_MC )_) Forman_NN1 and_CC Peniwati_NP1 ,_, 1998_MC )_) ._. 
</s>
<s>
In_II another_DD1 work_NN1 ,_, non-local_JJ patches_NN2 were_VBDR used_VVN to_TO train_VVI a_AT1 MRF_NP1 model_NN1 &lsqb;_( 9_MC &rsqb;_) ,_, which_DDQ preprocessed_VVD low_JJ quality_NN1 ,_, carbon_NN1 copy_NN1 document_NN1 images_NN2 ._. 
</s>
<s>
System-level_JJ metrics_NN1 presentations_NN2 should_VM be_VBI designed_VVN so_RR that_DD1 failed_JJ Design_NN1 Verification_NN1 dimensions_NN2 such_II21 as_II22 those_DD2 in_II Fig._NN1 14_MC would_VM drill_VVI down_RP into_II figures_NN2 like_II Fig._NN1 16_MC so_RR the_AT component_NN1 failure_NN1 impact_NN1 on_II system-level_JJ security_NN1 may_VM be_VBI analyzed_VVN and_CC understood_VVN ._. 
</s>
<s>
This_DD1 problem_NN1 is_VBZ broadly_RR described_VVN as_II trust_NN1 management_NN1 ._. 
</s>
<s>
NUTS_NN2 (_( Nomenclature_NN1 of_IO Territorial_JJ Units_NN2 for_IF Statistics_NN )_) is_VBZ a_AT1 geocode_NN1 standard_NN1 for_IF referencing_VVG the_AT subdivisions_NN2 of_IO the_AT UK_NP1 and_CC other_JJ EU_NP1 countries_NN2 for_IF statistical_JJ purposes_NN2 ,_, and_CC is_VBZ represented_VVN in_II DBpedia_NP1 ._. 
</s>
<s>
In_II the_AT information_NN1 diffusion_NN1 context_NN1 ,_, the_AT dissipation_NN1 is_VBZ related_VVN to_II the_AT loss_NN1 of_IO information_NN1 ._. 
</s>
<s>
TMS_NP1 relevance_NN1 is_VBZ defined_VVN as_II the_AT strength_NN1 of_IO the_AT relationship_NN1 between_II a_AT1 TMS_NP1 and_CC performance_NN1 on_II a_AT1 task_NN1 &lsqb;_( 17_MC &rsqb;_) ._. 
</s>
<s>
This_DD1 was_VBDZ done_VDN for_IF a_AT1 K-nearest_JJT neighbors_NN2 algorithm_NN1 called_VVN TCM-KNN_NP1 (_( Li_NP1 and_CC Guo_NP1 ,_, 2007_MC and_CC Li_NP1 et_RA21 al._RA22 ,_, 2007_MC )_) ._. 
</s>
<s>
InFig._NP1 5(E2)_FO ,_, we_PPIS2 can_VM see_VVI that_CST the_AT local_JJ egress_NN1 preference_NN1 (_( router_NN1 R1_FO )_) and_CC the_AT neighbor_NN1 '_NULL s_ZZ1 ingress_NN1 preference_NN1 (_( router_NN1 R3_FO )_) are_VBR violated_VVN ,_, without_IW the_AT knowledge_NN1 of_IO either_RR AS_CSA ._. 
</s>
<s>
However_RR ,_, C_ZZ1 provides_VVZ its_APPGE own_DA hazards_NN2 due_II21 to_II22 its_APPGE lack_NN1 of_IO advanced_JJ memory_NN1 management_NN1 ._. 
</s>
<s>
Classification_NN1 of_IO new_JJ samples_NN2 can_VM be_VBI performed_VVN after_II the_AT learning_NN1 of_IO the_AT intrinsic_JJ features_NN2 of_IO the_AT manifold_NN1 of_IO each_DD1 expression_NN1 ._. 
</s>
<s>
The_AT engine_NN1 blends_VVZ the_AT execution_NN1 of_IO UDFs_NP2 together_RL with_IW relational_JJ operators_NN2 using_VVG Just-In-Time_JJ tracing_NN1 compilation_NN1 techniques_NN2 ._. 
</s>
<s>
For_REX21 example_REX22 ,_, open_JJ question_NN1 answering_VVG over_RP unstructured_JJ documents_NN2 or_CC free_JJ text_NN1 has_VHZ been_VBN in_II the_AT focus_NN1 of_IO the_AT open-domain_JJ QA_NN1 track_NN1 introduced_VVN by_II TREC_NP1 from_II 1999_MC to_II 2007_MC ._. 
</s>
<s>
In_BCL21 order_BCL22 to_TO simplify_VVI the_AT process_NN1 of_IO model-building_NN1 ,_, we_PPIS2 separated_VVD the_AT model_NN1 into_II two_MC main_JJ components_NN2 :_: exercise_VV0 and_CC mental_JJ health_NN1 ._. 
</s>
<s>
This_DD1 means_VVZ that_CST the_AT economic_JJ value_NN1 of_IO an_AT1 additional_JJ unit_NN1 of_IO headwater_JJR supply_NN1 increases_VVZ as_CSA population_NN1 grows_VVZ ._. 
</s>
<s>
Ti_NP1 '_NULL s_ZZ1 lock_NN1 mode_NN1 on_II x_ZZ1 will_VM be_VBI upgraded_VVN to_II riR_NN1 (_( cell_NN1 &lsqb;_( 0_MC &rsqb;_) &lsqb;_( 1_MC1 &rsqb;_) )_) ,_, which_DDQ covers_VVZ both_RR lock_VV0 modes_NN2 rR_NNU and_CC iR_NN1 ._. 
</s>
<s>
Notice_VV0 that_CST the_AT conversion_NN1 from_II any_DD read_VV0 lock_NN1 to_II any_DD write_VV0 lock_NN1 results_NN2 in_II requesting_VVG a_AT1 write_VV0 lock_NN1 ._. 
</s>
<s>
The_AT SCM_NP1 transforms_VVZ the_AT observed_JJ FFPs_NP1 to_II the_AT converted_JJ FFPs_NP1 relatively_RR close_RR to_II the_AT ground-truth_JJ FFPs_NP1 ._. 
</s>
<s>
All_DB the_AT experiments_NN2 are_VBR executed_VVN using_VVG the_AT ICRGMM_NP1 algorithm_NN1 for_IF initial_JJ clustering_NN1 ._. 
</s>
<s>
The_AT tags_NN2 typically_RR consist_VV0 of_IO two_MC components_NN2 :_: standardised_JJ meta-data_NN ,_, such_II21 as_II22 artist_NN1 ,_, track_NN1 title_NN1 and_CC genre_NN1 ;_; and_CC free-text_NN1 that_CST users_NN2 provide_VV0 ._. 
</s>
<s>
We_PPIS2 applied_VVD the_AT same_DA algorithm_NN1 developed_VVN for_IF the_AT EVI_NN1 and_CC described_VVN in_II Section_NN1 2.2_MC ,_, to_II the_AT GPP_NP1 time_NNT1 series_NN to_TO derive_VVI start_NN1 and_CC end_NN1 of_IO episodes_NN2 timing_NN1 of_IO GPP_NP1 episodes_NN2 for_IF the_AT Howard_NP1 Springs_NN2 flux_NN1 site_NN1 ._. 
</s>
<s>
However_RR ,_, for_IF TP_NP1 and_CC TN_NP1 loads_NN2 ,_, Figs._NN2 ?_? 9_MC and_CC 10_MC ,_, it_PPH1 is_VBZ necessary_JJ to_TO have_VHI much_DA1 better_JJR fit_NN1 of_IO simulated_JJ data_NN with_IW observations_NN2 ,_, especially_RR for_IF TP_NN1 loads_NN2 ._. 
</s>
<s>
Therefore_RR ,_, the_AT algorithm_NN1 will_VM start_VVI putting_VVG 1_MC1 '_NULL s_ZZ1 in_II both_DB2 B(2)_FO and_CC B(3)_FO evenly_RR ,_, until_CS either_RR their_APPGE column_NN1 sums_NN2 reach_VV0 6k_FO or_CC B(3)_FO gets_VVZ filled_VVN ._. 
</s>
<s>
In_II order_NN1 to_TO further_RRR investigate_VVI the_AT nature_NN1 of_IO these_DD2 interactions_NN2 ,_, an_AT1 analysis_NN1 of_IO individual_JJ participant_NN1 interactions_NN2 would_VM be_VBI useful_JJ ._. 
</s>
<s>
Algorithms_NN2 for_IF automatically_RR transforming_VVG any_DD information_NN1 gathering_NN1 plan_NN1 into_II one_MC1 capable_JJ of_IO speculative_JJ execution_NN1 ._. 
</s>
<s>
Corresponding_VVG to_II the_AT colored_JJ clustering_NN1 results_NN2 ,_, Fig._NN1 6_MC ,_, Fig._NN1 7(a)_FO and_CC (_( b_ZZ1 )_) give_VV0 the_AT quantitative_JJ comparison_NN1 in_II31 terms_II32 of_II33 error_NN1 ratios_NN2 and_CC KL_NP1 distances_NN2 between_II the_AT estimated_JJ GMMs_NP1 and_CC the_AT standard_JJ GMMs_NP1 ,_, respectively_RR ._. 
</s>
<s>
MDM_NP1 is_VBZ typically_RR deployed_VVN through_II an_AT1 enrollment_NN1 or_CC provisioning_NN1 process_NN1 ._. 
</s>
<s>
For_IF itemc_NN1 ,_, a_AT1 1-itemset_JJ pattern_NN1 set_NN1 (_( c:4_FO )_) is_VBZ reported_VVN in_II the_AT first_MD step_NN1 ._. 
</s>
<s>
The_AT cost_NN1 function_NN1 reduction_NN1 is_VBZ relatively_RR insensitive_JJ to_II the_AT number_NN1 of_IO subintervals_NN2 for_IF the_AT Arenstorf_NN1 and_CC heat_NN1 equation_NN1 problems_NN2 ._. 
</s>
<s>
These_DD2 structures_NN2 subdivide_VV0 the_AT space_NN1 into_II regions_NN2 and_CC depending_II21 on_II22 the_AT algorithm_NN1 ,_, the_AT regions_NN2 are_VBR uniform_JJ or_CC non-uniform_NN1 ._. 
</s>
<s>
Although_CS ,_, until_II now_RT ,_, IT_NN1 researchers_NN2 have_VH0 used_VVN the_AT two_MC perspectives_NN2 independently_RR ,_, the_AT strategic_JJ alignment_NN1 model_NN1 proposed_VVN by_II Henderson_NP1 and_CC Venkatraman_NP1 (_( 1999_MC )_) can_VM be_VBI used_VVN ,_, from_II a_AT1 theoretical_JJ standpoint_NN1 ,_, to_TO integrate_VVI these_DD2 perspectives_NN2 ._. 
</s>
<s>
However_RR ,_, this_DD1 tidying_NN1 brings_VVZ problems_NN2 of_IO its_APPGE own_DA :_: it_PPH1 becomes_VVZ particularly_RR difficult_JJ to_TO ensure_VVI that_CST the_AT output_NN1 HTML_NP1 renders_VVZ exactly_RR like_II the_AT input_NN1 HTML_NP1 ._. 
</s>
<s>
Their_APPGE taxonomy_NN1 includes_VVZ security_NN1 assets_NN2 which_DDQ define_VV0 the_AT target_NN1 of_IO security_NN1 ,_, i.e._REX ,_, what_DDQ is_VBZ protected_VVN ,_, security_NN1 attributes_VVZ in_II the_AT form_NN1 of_IO ,_, security_NN1 threats_NN2 as_II the_AT causes_NN2 of_IO security_NN1 problems_NN2 ,_, security_NN1 solutions_NN2 as_CSA means_NN of_IO how_RRQ to_TO mitigate_VVI security_NN1 issues_NN2 ,_, and_CC security_NN1 metrics_NN2 to_TO measure_VVI the_AT effectiveness_NN1 of_IO security_NN1 technologies_NN2 ._. 
</s>
<s>
A_AT1 secondary_JJ goal_NN1 is_VBZ to_TO formalize_VVI these_DD2 constructs_NN2 ._. 
</s>
<s>
The_AT personomy_NN1 then_RT consists_VVZ of_IO all_DB photos_NN2 and_CC the_AT corresponding_JJ photo_NN1 sets_VVZ they_PPHS2 are_VBR assigned_VVN to_II ._. 
</s>
<s>
A_AT1 more_RGR detailed_JJ summary_NN1 is_VBZ provided_VVN inTable_JJ 4_NN1 (_( error_NN1 rates_NN2 are_VBR aggregated_VVN across_II different_JJ numbers_NN2 of_IO samples_NN2 for_IF each_DD1 dataset_NN1 andk_NN1 )_) ._. 
</s>
<s>
The_AT remainder_NN1 of_IO the_AT paper_NN1 is_VBZ organised_VVN as_CSA follows_VVZ :_: Section_NN1 2_MC examines_VVZ the_AT methodological_JJ status_NN1 of_IO longitudinal_JJ case_NN1 studies_NN2 in_II software_NN1 engineering_NN1 and_CC goes_VVZ onto_II define_VV0 the_AT longitudinal_JJ ,_, chronological_JJ case_NN1 study_NN1 ;_; Section_NN1 3_MC describes_VVZ the_AT design_NN1 of_IO the_AT particular_JJ LCCS_NP2 reported_VVD here_RL ;_; Section_NN1 4_MC provides_VVZ an_AT1 overview_NN1 to_II Project_NP1 C_NP1 ;_; Section_NN1 5_MC uses_VVZ qualitative_JJ and_CC structured_JJ qualitative_JJ data_NN to_TO describe_VVI the_AT socio-technical_JJ context_NN1 of_IO the_AT project_NN1 ;_; Section_NN1 6_MC describes_VVZ and_CC explores_VVZ the_AT behaviour_NN1 of_IO the_AT project_NN1 over_II time_NNT1 using_VVG multi-dimensional_JJ timelines_NN2 (_( MDTs_NP1 )_) ;_; Section_NN1 7_MC comments_NN2 on_II the_AT practical_JJ aspects_NN2 and_CC challenges_NN2 of_IO conducting_VVG LCCSs_NP1 ,_, and_CC on_II the_AT contribution_NN1 that_CST LCCSs_NP1 can_VM make_VVI ;_; finally_RR ,_, Section_NN1 8_MC considers_VVZ the_AT prospects_NN2 of_IO a_AT1 theory_NN1 of_IO project_NN1 behaviour_NN1 and_CC avenues_NN2 for_IF further_JJR research_NN1 ._. 
</s>
<s>
We_PPIS2 now_RT establish_VV0 the_AT relationship_NN1 between_II k-structures_NN2 and_CC j-structures_NN2 ,_, and_CC their_APPGE corresponding_JJ k_ZZ1 -structures_NN2 and_CC j_ZZ1 -structures_NN2 respectively_RR ._. 
</s>
<s>
To_TO account_VVI for_IF this_DD1 latter_DA ,_, an_AT1 activity-based_JJ transport_NN1 demand_NN1 model_NN1 was_VBDZ integrated_VVN in_II an_AT1 environmental_JJ modeling_NN1 framework_NN1 ._. 
</s>
<s>
The_AT program_NN1 can_VM also_RR directly_RR analyze_VVI media_NN directly_RR connected_VVN to_II the_AT analyst_NN1 '_NULL s_ZZ1 computerfor_NN1 example_NN1 ,_, with_IW a_AT1 write_VV0 blocker_NN1 ._. 
</s>
<s>
Messages_NN2 arise_VV0 along_II21 with_II22 particles_NN2 that_CST inform_VV0 of_IO the_AT illocutionary_JJ force_NN1 of_IO their_APPGE utterance_NN1 ._. 
</s>
<s>
We_PPIS2 also_RR investigated_VVD the_AT effect_NN1 of_IO constraints_NN2 over_II NFPs_NP1 on_II the_AT running_JJ time_NNT1 of_IO the_AT configuration_NN1 technique_NN1 proposed_VVN in_II Section_NN1 4_MC ._. 
</s>
<s>
In_II this_DD1 context_NN1 ,_, a_AT1 "_" unit_NN1 "_" included_VVD a_AT1 single_JJ HTML_NP1 file_NN1 ,_, a_AT1 source-code_JJ function_NN1 inside_II a_AT1 JavaScript_NN1 (_( JS_NP2 )_) ,_, JSP_NP1 or_CC PHP_NP1 file_NN1 ._. 
</s>
<s>
For_IF the_AT shock_NN1 reflection_NN1 problem_NN1 ,_, physically_RR the_AT reflected_JJ shock_NN1 remains_VVZ as_II a_AT1 shock_NN1 (_( a_AT1 compression_NN1 wave_NN1 )_) as_CSA it_PPH1 moves_VVZ away_II21 from_II22 the_AT shock_NN1 reflection_NN1 point_NN1 ._. 
</s>
<s>
We_PPIS2 note_VV0 that_CST γ_NULL =0_FO is_VBZ admissible_JJ ;_; in_II this_DD1 case_NN1 ,_, all_DB elements_NN2 are_VBR refined_JJ ._. 
</s>
<s>
While_CS we_PPIS2 currently_RR believe_VV0 that_CST this_DD1 problem_NN1 will_VM be_VBI intractable_JJ ,_, we_PPIS2 also_RR speculate_VV0 that_CST large_JJ and_CC useful_JJ subsets_NN2 of_IO this_DD1 problem_NN1 may_VM be_VBI tractable_JJ under_II some_DD reasonable_JJ assumptions_NN2 ._. 
</s>
<s>
All_DB requests_NN2 traverse_VV0 6_MC hop_NN1 paths_NN2 ,_, leading_VVG to_II a_AT1 completely_RR uniform_JJ traffic_NN1 distribution_NN1 ._. 
</s>
<s>
A_AT1 network_NN1 that_CST is_VBZ not_XX available_JJ ,_, i.e._REX ,_, with_IW low_JJ connectivity_NN1 ,_, has_VHZ a_AT1 slightly_RR higher_RRR predicted_VVN perceived_JJ risk_NN1 ,_, but_CCB the_AT difference_NN1 is_VBZ smaller_JJR than_CSN that_DD1 resulting_JJ from_II changing_JJ network_NN1 partitioning_NN1 or_CC wireless_NN1 ._. 
</s>
<s>
FV-RC_NP1 is_VBZ an_AT1 optimization_NN1 algorithm_NN1 with_IW a_AT1 huge_JJ search_NN1 space_NN1 ,_, which_DDQ makes_VVZ FV-RC_NP1 very_RG slow_JJ and_CC demands_VVZ a_AT1 lot_NN1 of_IO trials_NN2 with_IW different_JJ initial_JJ values_NN2 to_TO find_VVI a_AT1 good_JJ solution_NN1 ._. 
</s>
<s>
The_AT only_JJ consideration_NN1 was_VBDZ to_TO avoid_VVI migrations_NN2 into_II hijacked_JJ nodes_NN2 or_CC nodes_NN2 hosting_VVG bidders_NN2 or_CC the_AT AC_NN1 ._. 
</s>
<s>
However_RR ,_, it_PPH1 can_VM not_XX provide_VVI services_NN2 to_II others_NN2 as_II a_AT1 "_" server_NN1 "_" until_CS it_PPH1 has_VHZ become_VVN anS-node_NN1 ._. 
</s>
<s>
First_MD ,_, the_AT research_NN1 model_NN1 used_VVN expands_VVZ the_AT Shaft_NN1 and_CC of_IO program_NN1 comprehension_NN1 to_TO include_VVI concepts_NN2 related_VVN to_II the_AT unique_JJ nature_NN1 of_IO knowledge_NN1 work_NN1 ._. 
</s>
<s>
For_IF all_DB the_AT tests_NN2 we_PPIS2 keep_VV0 the_AT sample_NN1 size_NN1 same_DA and_CC equal_JJ to_II 1000_MC IPDs_NP2 ._. 
</s>
<s>
Looking_VVG at_II the_AT results_NN2 of_IO the_AT automated_JJ penetration_NN1 test_NN1 ,_, OpenEMR_NP1 had_VHD an_AT1 order_NN1 of_IO magnitude_NN1 more_RGR true_JJ positives_NN2 than_CSN Tolven_JJ eCHR_NN1 ._. 
</s>
<s>
The_AT Sinc-Collocation_NP1 approach_NN1 used_VVN in_II the_AT current_JJ paper_NN1 is_VBZ applied_VVN to_II the_AT complex_NN1 as_II31 well_II32 as_II33 the_AT real-value_JJ coupled_JJ systems_NN2 ._. 
</s>
<s>
Providing_VVG URIs_NN1 for_IF all_DB reviewers_NN2 and_CC reviewed_VVD things_NN2 gives_VVZ many_DA2 items_NN2 a_AT1 presence_NN1 on_II the_AT Semantic_JJ Web_NN1 which_DDQ they_PPHS2 would_VM not_XX have_VHI otherwise_RR ,_, and_CC enables_VVZ any_RR third_MD party_NN1 to_TO refer_VVI to_II these_DD2 items_NN2 in_II other_JJ RDF_NP1 statements_NN2 ._. 
</s>
<s>
Specifically_RR ,_, it_PPH1 would_VM be_VBI of_IO interest_NN1 to_TO collect_VVI traffic_NN1 traces_VVZ from_II a_AT1 large_JJ distributed_JJ CDN_NP1 and_CC empirically_RR study_VV0 the_AT bandwidth_NN1 cost_NN1 reduction_NN1 that_CST is_VBZ possible_JJ by_II using_VVG our_APPGE algorithmic_JJ ideas_NN2 ._. 
</s>
<s>
As_CSA norms_NN2 on_II air_NN1 quality_NN1 are_VBR often_RR described_VVN upon_II the_AT period_NN1 of_IO a_AT1 year_NNT1 ,_, being_VBG able_JK to_TO simulate_VVI a_AT1 complete_JJ year_NNT1 is_VBZ very_RG important_JJ and_CC using_VVG the_AT RIO-IFDM_JJ coupling_NN1 enables_VVZ this_DD1 ,_, although_CS a_AT1 validation_NN1 using_VVG temporal_JJ data_NN up_RG21 to_RG22 1_MC1 ?_? h_ZZ1 resolution_NN1 is_VBZ undertaken_VVN (_( Section_NN1 5_MC )_) ._. 
</s>
<s>
A_AT1 higher_JJR level_NN1 of_IO service_NN1 requires_VVZ an_AT1 increase_NN1 in_II the_AT network_NN1 capacity_NN1 to_TO meet_VVI the_AT minimum_JJ desirable_JJ pressures_NN2 of_IO the_AT network_NN1 ._. 
</s>
<s>
We_PPIS2 chose_VVD to_TO pay_VVI 0.03_MC USD_VV0 for_IF answering_VVG one_MC1 page_NN1 of_IO questions_NN2 ._. 
</s>
<s>
Realistically_RR ,_, the_AT window_NN1 size_NN1 must_VM be_VBI set_VVN to_II some_DD value_NN1 larger_JJR then_RT the_AT longest_JJT activity_NN1 length_NN1 ,_, taking_VVG into_II account_NN1 a_AT1 feasible_JJ increase_NN1 in_II possible_JJ activity_NN1 duration_NN1 ._. 
</s>
<s>
A_AT1 user_NN1 can_VM access_VVI timetable_NN1 information_NN1 at_II a_AT1 particular_JJ stop_NN1 by_II tapping_VVG on_II it_PPH1 ._. 
</s>
<s>
Failure_NN1 to_TO conform_VVI to_II standards_NN2 :_: On_II the_AT web_NN1 ,_, exceptions_NN2 are_VBR the_AT rule_NN1 ._. 
</s>
<s>
This_DD1 section_NN1 provides_VVZ an_AT1 example_NN1 of_IO the_AT effect_NN1 of_IO a_AT1 random_JJ field_NN1 on_II the_AT critical_JJ load_NN1 factor_NN1 of_IO an_AT1 arch_JJ structure_NN1 ._. 
</s>
<s>
This_DD1 lemma_NN1 is_VBZ easily_RR proved_VVN using_VVG a_AT1 Taylor_NP1 expansion_NN1 of_IO uR_FO and_CC ._. 
</s>
<s>
Its_APPGE audit_NN1 and_CC strategy_NN1 were_VBDR produced_VVN by_II small_JJ team_NN1 ,_, led_VVN by_II the_AT City_NN1 '_NULL s_ZZ1 community_NN1 safety_NN1 manager_NN1 ._. 
</s>
<s>
FND_NP1 takes_VVZ care_NN1 of_IO two_MC tasks_NN2 ,_, namely_REX ,_, ?_? 
</s>
<s>
Fault_NN1 Notification_NN1 Message_NN1 Generation_NN1 ,_, and_CC Fault_NN1 Notification_NN1 Message_NN1 Forwarding_NN1 ._. 
</s>
<s>
Finally_RR ,_, with_IW a_AT1 small_JJ number_NN1 of_IO popular_JJ internal_JJ nodes_NN2 uT_VV0 we_PPIS2 associate_VV0 a_AT1 measurement_NN1 (_( mua_NN1 ,_, mus_NN2 )_) &lsqb;_( 0,1_MC &rsqb;_) 2_MC ._. 
</s>
<s>
However_RR ,_, in_II practice_NN1 ,_, it_PPH1 is_VBZ very_RG difficult_JJ to_TO get_VVI this_DD1 switch_NN1 over_II correct_JJ ._. 
</s>
<s>
Note_VV0 that_CST we_PPIS2 also_RR assume_VV0 that_CST the_AT boundary_NN1 and_CC initial_JJ data_NN ,_, Eqs._NP1 (_( 5_MC )_) ,_, (_( 6_MC )_) and(7)_FO ,_, can_VM be_VBI accurately_RR approximated_VVN at_II the_AT grid-scale_JJ ._. 
</s>
<s>
It_PPH1 is_VBZ also_RR responsible_JJ for_IF interfacing_NN1 with_IW the_AT Database_NN1 Manager_NN1 and_CC sending_VVG database_NN1 read/write_VV0 commands_NN2 based_VVN on_II SSXD_NP1 ._. 
</s>
<s>
Denote_VV0 the_AT set_NN1 of_IO coalitions_NN2 which_DDQ have_VH0 a_AT1 value_NN1 of_IO ui_NNU as_II Cui_NP1 (_( so_RR if_CS C∈Cui_FO thenv(C)=ui_FO )_) ._. 
</s>
<s>
The_AT constructs_NN2 are_VBR presented_VVN within_II the_AT grid_NN1 as_CSA bipolar_JJ distinctions_NN2 ;_; this_DD1 is_VBZ important_JJ as_CSA Kelly_NP1 argues_VVZ that_CST people_NN tend_VV0 to_TO evaluate_VVI in_II31 terms_II32 of_II33 likeness_NN1 and_CC difference_NN1 ,_, hence_RR depicting_VVG constructs_NN2 on_II a_AT1 bipolar_JJ continuum_NN1 allows_VVZ one_PN1 to_TO identify_VVI how_RRQ elements_NN2 might_VM differ_VVI in_II fundamental_JJ ways_NN2 ._. 
</s>
<s>
The_AT notion_NN1 of_IO state_NN1 in_II RESTful_JJ applications_NN2 might_VM cause_VVI confusion_NN1 ._. 
</s>
<s>
As_CSA stated_VVN earlier_RRR (_( Section_NN1 5.2_MC )_) ,_, one_MC1 of_IO the_AT key_JJ advantages_NN2 of_IO a_AT1 1+n_FO strategy_NN1 is_VBZ that_CST the_AT user_NN1 does_VDZ not_XX have_VHI to_TO score_VVI all_DB the_AT solutionsjust_NN1 the_AT ones_NN2 that_CST are_VBR the_AT bestthis_NN1 will_VM greatly_RR enhance_VVI the_AT level_NN1 of_IO engagement_NN1 with_IW the_AT process_NN1 ._. 
</s>
<s>
InTables_NP1 2_MC and_CC 3_MC ,_, this_DD1 is_VBZ called_VVN Bad_JJ node_NN1 with_IW Bad_JJ keys_NN2 ._. 
</s>
<s>
To_TO validate_VVI the_AT functionality_NN1 ,_, we_PPIS2 simulated_VVD a_AT1 simple_JJ use-case_JJ scenario_NN1 (_( Fig._NN1 19_MC )_) over_II NS3&lsqb;20&rsqb;_FO showing_VVG distributed_JJ application_NN1 deployment_NN1 over_II multi-cloud_JJ environment_NN1 ._. 
</s>
<s>
We_PPIS2 fix_VV0 some_DD fraction_NN1 f_ZZ1 ,_, and_CC for_IF each_DD1 internal_JJ node_NN1 uT_NN1 with_IW vector_NN1 ,_, we_PPIS2 keep_VV0 only_RR the_AT largest_JJT max_NN1 entries_NN2 of_IO ._. 
</s>
<s>
We_PPIS2 used_VVD the_AT AS_CSA information_NN1 obtained_VVN above_RL to_TO characterize_VVI the_AT transit_NN1 policy_NN1 violations_NN2 in_II our_APPGE case_NN1 study_NN1 ._. 
</s>
<s>
Determination_NN1 of_IO the_AT best_JJT value_NN1 for_IF this_DD1 combination_NN1 coefficient_NN1 is_VBZ required_VVN to_TO produce_VVI the_AT most_RGT outstanding_JJ detection_NN1 results_NN2 ._. 
</s>
<s>
The_AT only_JJ component_NN1 done_VDN on_II a_AT1 per_II query_NN1 basis_NN1 is_VBZ the_AT utilizing_VVG the_AT matrix_NN1 elements_NN2 for_IF the_AT query_NN1 modification_NN1 ._. 
</s>
<s>
From_II the_AT chosen_JJ representations_NN2 ,_, the_AT final_JJ query_NN1 graph_NN1 is_VBZ generated_VVN ._. 
</s>
<s>
Following_VVG this_DD1 assumption_NN1 ,_, Roweis_NP1 and_CC Saul_NP1 &lsqb;_( 15_MC &rsqb;_) proposed_VVD to_TO characterize_VVI the_AT local_JJ geometry_NN1 in_II the_AT neighborhood_NN1 of_IO each_DD1 data_NN point_NN1 by_II coefficients_NN2 that_CST linearly_RR reconstruct_VV0 the_AT data_NN point_NN1 from_II its_APPGE neighbors_NN2 ,_, and_CC they_PPHS2 gave_VVD an_AT1 algorithm_NN1 to_TO pursue_VVI these_DD2 linear_JJ reconstruction_NN1 coefficients_NN2 ._. 
</s>
<s>
In_II such_DA a_AT1 graph_NN1 a_AT1 cycle_NN1 (_( i.e._REX a_AT1 closed_JJ contour_NN1 )_) can_VM now_RT be_VBI defined_VVN as_II a_AT1 path_NN1 that_CST starts_VVZ in_II a_AT1 source_NN1 vertex_NN1 and_CC ends_VVZ in_II the_AT corresponding_JJ goal_NN1 vertex_NN1 ._. 
</s>
<s>
Taking_VVG an_AT1 example_NN1 TPR_NP1 of_IO 0.7_MC the_AT best_JJT detectors_NN2 are_VBR ,_, in_II order_NN1 of_IO increasing_JJ performance_NN1 ,_, convolution_NN1 (_( FPR_NP1 :_: 0.246_MC )_) ;_; PCA_NP1 (_( FPR_NP1 :_: 0.213_MC )_) ,_, bar_NN1 multi-scale_NN1 (_( FPR_NP1 :_: 0.134_MC )_) ,_, and_CC bar_NN1 fixed-scale_NN1 (_( FPR_NP1 :_: 0.102_MC )_) ._. 
</s>
<s>
Also_RR the_AT research_NN1 community_NN1 has_VHZ no_AT common_JJ view_NN1 on_II the_AT security_NN1 properties_NN2 that_CST a_AT1 CMS_NP1 has_VHZ to_TO provide_VVI ._. 
</s>
<s>
Investigating_VVG any_DD research_NN1 question_NN1 requires_VVZ that_CST the_AT methodology_NN1 uncover_VV0 the_AT multiple_JJ connections_NN2 that_CST typically_RR exist_VV0 between_II the_AT various_JJ parts_NN2 of_IO a_AT1 project_NN1 ._. 
</s>
<s>
Here_RL (_( a_ZZ1 )_) is_VBZ a_AT1 case_NN1 ,_, in_II which_DDQ is_VBZ relatively_RR large_JJ and_CC Rout_NN1 is_VBZ small_JJ ,_, meaning_VVG that_DD1 numerical_JJ vectors_NN2 are_VBR not_XX easily_RR clustered_VVN but_CCB network_NN1 nodes_NN2 can_VM be_VBI clustered_VVN easily_RR ._. 
</s>
<s>
The_AT union_NN1 of_IO these_DD2 sets_NN2 is_VBZ abbreviated_VVN as_CSA IBL_NP1 ._. 
</s>
<s>
Finally_RR ,_, given_VVN the_AT entity_NN1 URI_NN2 we_PPIS2 retrieve_VV0 all_DB its_APPGE types_NN2 (_( from_II a_AT1 background_NN1 RDF_NP1 corpus_NN1 or_CC from_II a_AT1 previously_RR created_VVN inverted_JJ index_NN1 )_) and_CC rank_VV0 them_PPHO2 given_VVN a_AT1 context_NN1 ._. 
</s>
<s>
The_AT following_JJ is_VBZ thereby_RR a_AT1 more_RGR specific_JJ version_NN1 of_IO Principle_NN1 3_MC ._. 
</s>
<s>
In_II some_DD cases_NN2 (_( e.g._REX ,_, TTL-limited_JJ flooding_NN1 )_) coverage_NN1 can_VM be_VBI traded_VVN off_RP with_IW network_NN1 efficiency_NN1 ._. 
</s>
<s>
Certainly_RR comparison_NN1 websites_NN2 offer_VV0 access_NN1 to_II a_AT1 wider_JJR market_NN1 for_IF niche_NN1 players_NN2 and_CC Anderson_NP1 actually_RR refers_VVZ to_II them_PPHO2 in_II his_APPGE work_NN1 ._. 
</s>
<s>
Examples_NN2 include_VV0 age_NN1 ,_, gender_NN1 ,_, height_NN1 ,_, weight_NN1 ,_, ethnicity_NN1 and_CC distinctive_JJ markings_NN2 (_( scars_VVZ ,_, marks_NN2 and_CC tattoos_NN2 )_) ._. 
</s>
<s>
We_PPIS2 have_VH0 noted_VVN that_DD1 DUMMY_NN1 ,_, both_RR in_II TCSG-T_NP1 and_CC in_II WTSG-T_NP1 ,_, is_VBZ in_II co-NP_NP1 ;_; it_PPH1 remains_VVZ to_TO show_VVI that_CST it_PPH1 is_VBZ co-NP_NP1 hard_RR ._. 
</s>
<s>
As_CSA Section_NN1 6.5_MC points_VVZ out_RP ,_, this_DD1 understanding_NN1 also_RR helped_VVD us_PPIO2 explain_VVI some_DD of_IO our_APPGE findings_NN2 ._. 
</s>
<s>
Query_VV0 syntax_NN1 can_VM be_VBI saved_VVN and_CC modified_VVN subsequently_RR such_CS21 that_CS22 new_JJ users_NN2 get_VV0 the_AT chance_NN1 to_TO learn_VVI more_RGR expressive_JJ queries_NN2 used_VVN by_II other_JJ researchers_NN2 ._. 
</s>
<s>
That_CST the_AT IISP_NN1 and_CC BCS_NP2 both_RR run_VV0 security_NN1 certification_NN1 schemes_NN2 (_( and_CC that_CST the_AT IISP_NN1 '_NULL s_ZZ1 framework_NN1 has_VHZ apparently_RR found_VVN favour_NN1 with_IW Government_NN1 with_IW regards_VVZ to_II assessing_VVG education_NN1 )_) is_VBZ arguably_RR an_AT1 example_NN1 of_IO (_( constructive_JJ )_) Abbot-type_JJ splinter_NN1 competition_NN1 for_IF control_NN1 of_IO a_AT1 body_NN1 of_IO knowledge_NN1 ,_, along_II21 with_II22 the_AT partial_JJ intervention_NN1 of_IO the_AT state_NN1 ._. 
</s>
<s>
This_DD1 is_VBZ reasonable_JJ because_CS a_AT1 larger_JJR threshold_NN1 percentage_NN1 might_VM cause_VVI small_JJ but_CCB genuine_JJ clusters_NN2 to_TO be_VBI mistakenly_RR removed_VVN from_II the_AT final_JJ solution_NN1 ._. 
</s>
<s>
The_AT value_NN1 of_IO this_DD1 of_RR21 course_RR22 depends_VVZ on_II whether_CSW the_AT cospectral_JJ pairs_NN2 of_IO graphs_NN2 are_VBR the_AT same_DA in_II each_DD1 spectrum_NN1 ._. 
</s>
<s>
We_PPIS2 conclude_VV0 that_CST there_EX is_VBZ little_RR likely_JJ benefit_NN1 to_II further_JJR exploration_NN1 of_IO the_AT kind_NN1 of_IO general_JJ concept_NN1 tagging_VVG used_JJ in_II TREC_NP1 Genomics_NP1 and_CC elsewhere_RL ._. 
</s>
<s>
Full_JJ Text_NN1 via_II CrossRef_NP1 |_NULL View_NN1 Record_NN1 in_II Scopus_NP1 |_NULL Cited_VVD By_II in_II Scopus_NP1 (_( 144_MC )_) &lsqb;_( Scandura_NP1 and_CC Williams_NP1 ,_, 2000_MC &rsqb;_) )_) ._. 
</s>
<s>
This_DD1 paper_NN1 provides_VVZ interesting_JJ comparative_JJ results_NN2 allowing_VVG researchers_NN2 to_TO get_VVI a_AT1 clear_JJ insight_NN1 into_II the_AT best_JJT framework_NN1 for_IF fusion_NN1 ._. 
</s>
<s>
Fortran_NP1 has_VHZ traditionally_RR been_VBN the_AT language_NN1 of_IO choice_NN1 in_II the_AT scientific_JJ computing_NN1 domain_NN1 and_CC a_AT1 large_JJ portion_NN1 of_IO legacy_NN1 scientific_JJ software_NN1 written_VVN in_II Fortran_NP1 is_VBZ still_RR used_VVN to_TO build_VVI modern_JJ scientific_JJ applications_NN2 ._. 
</s>
<s>
Using_VVG the_AT LPGF_NP1 at_II different_JJ scales_NN2 represents_VVZ the_AT inner_JJ and_CC central_JJ part_NN1 of_IO an_AT1 object_NN1 more_RRR than_CSN the_AT boundary_NN1 ._. 
</s>
<s>
In_II many_DA2 settings_NN2 traditional_JJ generative_JJ or_CC discriminative_JJ methods_NN2 either_RR are_VBR infeasible_JJ or_CC fail_VV0 to_TO provide_VVI acceptable_JJ results_NN2 and_CC generalization_NN1 to_II new_JJ data_NN ._. 
</s>
<s>
For_IF this_DD1 reason_NN1 ,_, this_DD1 policy_NN1 is_VBZ named_VVN partially_RR pessimistic_JJ ._. 
</s>
<s>
Similarly_RR to_II that_DD1 work_NN1 ,_, we_PPIS2 build_VV0 our_APPGE approaches_NN2 using_VVG large_JJ knowledge_NN1 graphs_NN2 such_II21 as_II22 YAGO_NP1 and_CC DBpedia_NP1 ._. 
</s>
<s>
Each_DD1 paradigm_NN1 comes_VVZ with_IW its_APPGE own_DA strengths_NN2 and_CC weaknesses_NN2 ,_, and_CC as_II a_AT1 result_NN1 is_VBZ better_RRR suited_VVN to_II answering_VVG particular_JJ research_NN1 questions_NN2 ._. 
</s>
<s>
Scientists_NN2 may_VM need_VVI to_TO extend_VVI the_AT dataset_NN1 with_IW their_APPGE own_DA knowledge_NN1 ._. 
</s>
<s>
The_AT first_MD issue_NN1 is_VBZ the_AT securing_NN1 of_IO channels_NN2 between_II two_MC communicating_JJ parties_NN2 to_TO avoid_VVI eavesdropping_NN1 ,_, tampering_VVG with_IW the_AT transmission_NN1 or_CC session_NNT1 hijacking_VVG ._. 
</s>
<s>
Similar_JJ was_VBDZ the_AT case_NN1 of_IO the_AT Journalist_NN1 class_NN1 ,_, but_CCB with_IW ,_, respectively_RR ,_, 11_MC and_CC 5_MC out_II21 of_II22 16_MC axioms_NN2 ._. 
</s>
<s>
Fig._NN1 12depicts_FO the_AT steps_NN2 of_IO performing_VVG ._. 
</s>
<s>
One_MC1 of_IO the_AT multifaceted_JJ influences_NN2 with_IW both_DB2 external_JJ and_CC internal_JJ dimensions_NN2 relates_VVZ to_II security_NN1 threats_NN2 ._. 
</s>
<s>
This_DD1 approach_NN1 attempts_NN2 to_TO bridge_VVI the_AT gap_NN1 between_II information_NN1 systems_NN2 (_( IS_VBZ )_) staff_NN and_CC management_NN1 within_II the_AT security_NN1 policy_NN1 design_NN1 process_NN1 ._. 
</s>
<s>
The_AT ?_? h_ZZ1 ?_? payloads_NN2 of_IO original_JJ packets_NN2 are_VBR XORed_JJ together_RL to_TO form_VVI a_AT1 unique_JJ payload_NN1 calledtemp_NN1 ?_? in_BCL21 order_BCL22 to_TO reduce_VVI the_AT computation_NN1 complexity_NN1 ._. 
</s>
<s>
Once_CS SCD_NP1 features_NN2 (_( such_II21 as_II22 length_NN1 and_CC entropy_NN1 )_) were_VBDR derived_VVN from_II string_NN1 features_NN2 ,_, we_PPIS2 filtered_VVD these_DD2 original_JJ string_NN1 features_NN2 from_II the_AT dataset_NN1 since_CS we_PPIS2 require_VV0 only_RR numeric_JJ features_NN2 for_IF our_APPGE classifier_NN1 ._. 
</s>
<s>
As_II a_AT1 drawback_NN1 ,_, this_DD1 approach_NN1 might_VM not_XX be_VBI as_RG intuitive_JJ as_CSA the_AT cost_NN1 function_NN1 JA_NP1 ._. 
</s>
<s>
In_II an_AT1 equation_NN1 such_DA asP(Q>x)=y_FO ,_, we_PPIS2 shall_VM refer_VVI toQas_NN2 the_AT stationary_JJ buffer_NN1 level_NN1 or_CC queue_VV0 size_NN1 andxas_NN2 the_AT threshold_NN1 thatQexceeds_NN2 ._. 
</s>
<s>
If_CS the_AT error_NN1 for_IF every_AT1 iSet_NN1 is_VBZ below_II the_AT error_NN1 threshold_NN1 ,_, Qth_NP1 ,_, the_AT aggregated_JJ error_NN1 across_II all_DB the_AT iSets_NN2 is_VBZ computed_VVN and_CC saved_VVN ._. 
</s>
<s>
PRIDE_NN1 proposes_VVZ to_TO use_VVI the_AT security_NN1 administrator_NN1 '_NULL s_ZZ1 knowledge_NN1 about_II WMN_NP1 traffic_NN1 to_TO distribute_VVI IDS_NN2 functions_NN2 (_( i.e._REX ,_, Snort_VV0 detection_NN1 rules_NN2 )_) to_II the_AT nodes_NN2 along_II WMN_NP1 routing_NN1 paths_NN2 ._. 
</s>
<s>
It_PPH1 assigns_VVZ the_AT largest_JJT unallocated_JJ score_NN1 to_II the_AT largest_JJT gap_NN1 ._. 
</s>
<s>
Similar_JJ methods_NN2 include_VV0 those_DD2 described_VVN in_II &lsqb;_( 14_MC &rsqb;_) ,_, &lsqb;_( 8_MC &rsqb;_) ,_, &lsqb;_( 9_MC &rsqb;_) ,_, &lsqb;_( 10_MC &rsqb;_) ,_, &lsqb;_( 29_MC &rsqb;_) ,_, in_II which_DDQ tensors_NN2 (_( regular_JJ hypergraphs_NN2 )_) are_VBR used_VVN to_TO represent_VVI the_AT multiple_JJ relationships_NN2 between_II objects_NN2 ._. 
</s>
<s>
Dozens_NNO2 of_IO papers_NN2 were_VBDR a_AT1 direct_JJ result_NN1 of_IO the_AT Mothra_NP1 project_NN1 ,_, and_CC hundreds_NNO2 of_IO papers_NN2 were_VBDR written_VVN after_II the_AT project_NN1 ,_, but_CCB based_VVN on_II the_AT product_NN1 ._. 
</s>
<s>
We_PPIS2 have_VH0 tested_VVN our_APPGE Fractal_JJ Ravens_NN2 algorithm_NN1 on_II all_DB problems_NN2 associated_VVN with_IW the_AT four_MC main_JJ variations_NN2 of_IO the_AT Raven_NN1 '_NULL s_ZZ1 Progressive_JJ Matrices_NN2 Tests_NN2 :_: 60_MC problems_NN2 of_IO the_AT Standard_JJ Progressive_JJ Matrices_NN2 test_VV0 ,_, 48_MC problems_NN2 of_IO the_AT Advanced_JJ Progressive_JJ Matrices_NN2 test_VV0 ,_, 36_MC problems_NN2 of_IO the_AT Coloured_JJ Progressive_JJ Matrices_NN2 test_VV0 ,_, and_CC 60_MC problems_NN2 of_IO the_AT SPM_NP1 Plus_II test_NN1 ._. 
</s>
<s>
The_AT ant_NN1 and_CC data_NN routing_VVG probabilities_NN2 are_VBR generated_VVN according_II21 to_II22 different_JJ functions_NN2 of_IO the_AT trip_NN1 times_NNT2 estimates_VVZ ,_, to_TO be_VBI described_VVN shortly_RR ._. 
</s>
<s>
Two_MC major_JJ issues_NN2 make_VV0 the_AT dialog_NN1 modeling_NN1 problem_NN1 hard_RR to_TO manage_VVI ,_, and_CC these_DD2 concern_VVI the_AT kind_NN1 of_IO features_NN2 to_TO be_VBI extracted_VVN and_CC the_AT mathematical_JJ model_NN1 to_TO be_VBI applied_VVN to_TO process_VVI such_DA features_NN2 in_BCL21 order_BCL22 to_TO realize_VVI the_AT nature_NN1 of_IO the_AT dialog_NN1 ._. 
</s>
<s>
In_II an_AT1 XML_NP1 tree_NN1 ,_, the_AT leaf_NN1 nodes_NN2 represent_VV0 the_AT content_NN1 ._. 
</s>
<s>
This_DD1 has_VHZ led_VVN to_II a_AT1 demand_NN1 for_IF stable_JJ ,_, governed_VVD data_NN standards_NN2 &lsqb;_( 11_MC &rsqb;_) ,_, &lsqb;_( 12_MC &rsqb;_) ,_, which_DDQ are_VBR "_" technical_JJ documents_NN2 designed_VVN to_TO be_VBI used_VVN as_II a_AT1 rule_NN1 ,_, guideline_NN1 or_CC definition_NN1 ._. 
</s>
<s>
The_AT current_JJ prototype_NN1 lacks_VVZ the_AT logic_NN1 required_VVN to_TO safely_RR reallocate_VVI the_AT work_NN1 array_NN1 to_II a_AT1 different_JJ address_NN1 without_IW locks_NN2 ,_, but_CCB this_DD1 will_VM be_VBI added_VVN in_II the_AT near_JJ future_NN1 ._. 
</s>
<s>
The_AT NRS_NP1 evidence_NN1 must_VM be_VBI signed_VVN by_II the_AT sender_NN1 '_NULL s_ZZ1 MTA_NN1 with_IW a_AT1 QES_NP1 ._. 
</s>
<s>
We_PPIS2 run_VV0 the_AT experiment_NN1 with_IW two_MC separate_JJ pairs_NN2 of_IO servers_NN2 ,_, Black_JJ and_CC Gray_NP1 ,_, that_CST correspond_VV0 to_II the_AT black-_NN1 and_CC gray-box_JJ approaches_NN2 ,_, respectively_RR ._. 
</s>
<s>
Since_CS we_PPIS2 only_RR need_VV0 to_TO identify_VVI the_AT permissions_NN2 that_CST u_ZZ1 will_VM be_VBI able_JK to_TO infer_VVI if_CSW he_PPHS1 activated_VVD R_ZZ1 ,_, in_II line_NN1 6_MC ,_, Q_ZZ1 is_VBZ initialized_VVN to_TO contain_VVI only_RR newly_RR inferred_VVN information_NN1 ._. 
</s>
<s>
In_II APLS_NP2 ,_, the_AT two_MC steps_NN2 are_VBR connected_VVN through_II an_AT1 application_NN1 meta-tag_NN1 (_( or_CC simply_RR referred_VVN to_II as_II the_AT meta-tag_NN1 )_) that_CST encodes_VVZ the_AT result_NN1 of_IO the_AT application-level_JJ classification_NN1 ._. 
</s>
<s>
The_AT discontinuous_JJ and_CC singular_JJ quadratures_NN2 were_VBDR used_VVN in_II the_AT benchmark_NN1 problem_NN1 of_IO an_AT1 inclined_JJ crack_NN1 under_II biaxial_JJ tensionthe_NN1 mixed-mode_NN1 stress_NN1 intensity_NN1 factors_NN2 calculated_VVD using_VVG the_AT X-FEM_NN1 were_VBDR in_II good_JJ agreement_NN1 with_IW the_AT exact_JJ solution_NN1 ,_, which_DDQ provided_VVD further_JJR validation_NN1 on_II the_AT accuracy_NN1 of_IO the_AT integration_NN1 schemes_NN2 ._. 
</s>
<s>
P-CSCF_NP1 can_VM also_RR act_VVI as_II a_AT1 User_NN1 Agent_NN1 on_II31 behalf_II32 of_II33 the_AT actual_JJ Client_NN1 User_NN1 Agent_NN1 ._. 
</s>
<s>
The_AT MAD#_FO user_NN1 can_VM define_VVI different_JJ subsets_NN2 of_IO the_AT Type-B_NP1 data_NN vector_NN1 to_TO evaluate_VVI the_AT likelihood_NN1 ._. 
</s>
<s>
A_AT1 second_MD set_NN1 of_IO adaptivity_NN1 scenarios_NN2 is_VBZ presented_VVN in_II Section_NN1 5_MC ,_, based_VVN on_II a_AT1 cell_NN1 biology_NN1 simulation_NN1 code_NN1 ._. 
</s>
<s>
From_II an_AT1 application_NN1 engineering_NN1 perspective_NN1 ,_, the_AT basic_JJ reusability_NN1 strategy_NN1 in_II the_AT software_NN1 product_NN1 lines_NN2 is_VBZ based_VVN on_II the_AT reutilization_NN1 of_IO two_MC principal_JJ assets_NN2 :_: the_AT software_NN1 architecture_NN1 of_IO the_AT line_NN1 (_( SPLA_NP1 )_) that_CST defines_VVZ the_AT family_NN1 of_IO products_NN2 that_CST belong_VV0 to_II the_AT SPL_NP1 ,_, and_CC the_AT software_NN1 artifacts_NN2 that_CST provide_VV0 the_AT architecture_NN1 to_TO build_VVI a_AT1 new_JJ product_NN1 &lsqb;_( 6_MC &rsqb;_) ._. 
</s>
<s>
This_DD1 paper_NN1 is_VBZ organised_VVN as_CSA follows_VVZ ._. 
</s>
<s>
In_II Ariu_NP1 et_RA21 al_RA22 ._. 
</s>
<s>
(_( 2011_MC )_) ,_, various_JJ combinations_NN2 of_IO the_AT HMM_UH scores_NN2 are_VBR tested_VVN ._. 
</s>
<s>
It_PPH1 can_VM be_VBI observed_VVN that_CST a_AT1 45%_NNU reduction_NN1 of_IO computational_JJ time_NNT1 was_VBDZ obtained_VVN for_IF the_AT first_MD two_MC selections_NN2 ._. 
</s>
<s>
Temporal_JJ variability_NN1 :_: mean_VV0 absolute_JJ bias_NN1 ,_, RMSE_NP1 ,_, R2_FO ,_, bias_NN1 corrected_VVD RMSE_NP1 on_II an_AT1 hourly_JJ averaged_JJ time_NNT1 series_NN over_II all_DB stations_NN2 ._. 
</s>
<s>
Most_DAT of_IO the_AT lessons_NN2 to_TO learn_VVI from_II Java_NP1 '_NULL s_ZZ1 vulnerabilities_NN2 echo_VV0 Saltzer_NP1 and_CC Schroeder_JJR '_NULL s_ZZ1 classic_JJ principles_NN2 ,_, especially_RR economy_NN1 of_IO mechanism_NN1 ,_, least_DAT privilege_NN1 and_CC fail-safe_JJ defaults_NN2 ._. 
</s>
<s>
The_AT results_NN2 of_IO this_DD1 experiment_NN1 can_VM ,_, for_REX21 example_REX22 ,_, be_VBI used_VVN to_TO determine_VVI the_AT amino_NN1 acid_NN1 groupings_NN2 that_CST maximise_VV0 compressibility_NN1 ._. 
</s>
<s>
This_DD1 distinction_NN1 at_II the_AT semantic_JJ layer_NN1 is_VBZ dependent_JJ upon_II the_AT ontology_NN1 formalism_NN1 employed_VVD ._. 
</s>
<s>
For_IF all_DB entailments_NN2 ,_, justifications_NN2 can_VM be_VBI computed_VVN on_II demand_NN1 ._. 
</s>
<s>
These_DD2 networks_NN2 had_VHD topologies_NN2 that_CST were_VBDR closer_JJR to_II the_AT likely_JJ topologies_NN2 of_IO real_JJ networks_NN2 than_CSN the_AT idealized_JJ structures_NN2 used_VVN in_II the_AT analysis_NN1 of_IO Section4_FO ._. 
</s>
<s>
The_AT size_NN1 of_IO a_AT1 target_NN1 model_NN1 equals_VVZ the_AT number_NN1 of_IO system_NN1 variables_NN2 (_( including_II hidden_JJ variables_NN2 )_) in_II the_AT model_NN1 ._. 
</s>
<s>
These_DD2 are_VBR essentially_RR learned_VVN when_CS the_AT models_NN2 meet_VV0 many_DA2 training_NN1 examples_NN2 in_II which_DDQ such_DA a_AT1 token_NN1 is_VBZ part_NN1 of_IO the_AT text_NN1 associated_VVN with_IW a_AT1 similar_JJ pattern_NN1 of_IO input_NN1 triples_NN2 ._. 
</s>
<s>
So_RR after_CS all_DB pairs_NN2 of_IO similar_JJ requirements_NN2 are_VBR processed_VVN ,_, the_AT number_NN1 from_II each_DD1 of_IO these_DD2 ordered_JJ pair_NN give_VV0 us_PPIO2 the_AT occurrence_NN1 count_NN1 of_IO the_AT verb_NN1 synonym_NN1 pair_NN within_II the_AT whole_JJ requirement_NN1 document_NN1 ._. 
</s>
<s>
Here_RL the_AT SPs_NN2 constitute_VV0 95%_NNU of_IO the_AT total_JJ population_NN1 while_CS there_EX are_VBR no_AT LRPs_NP1 or_CC HRPs_NP1 in_II the_AT population_NN1 ._. 
</s>
<s>
Practitioners_NN2 should_VM also_RR consider_VVI using_VVG companies_NN2 such_II21 as_II22 Bazaarvoice.com_NP1 that_CST searches_VVZ the_AT Internet_NN1 for_IF defamatory_JJ ,_, racist_NN1 ,_, or_CC other_JJ such_DA language_NN1 posted_VVN about_II their_APPGE products_NN2 or_CC services_NN2 ._. 
</s>
<s>
Automatic_JJ feature_NN1 selection_NN1 was_VBDZ shown_VVN to_TO produce_VVI best_RRT overall_JJ accuracy_NN1 ._. 
</s>
<s>
This_DD1 comparison_NN1 is_VBZ done_VDN in_II Fig._NN1 6_MC ._. 
</s>
<s>
During_II step_NN1 4a_FO ,_, the_AT algorithm_NN1 creates_VVZ a_AT1 SensitivityPoint_NN1 for_IF each_DD1 SensitivityPoint_NN1 concept_NN1 declared_VVN inside_II the_AT of_IO the_AT SBM_NP1 ._. 
</s>
<s>
The_AT sliding_JJ window_NN1 approach_NN1 is_VBZ best_RRT described_VVN by_II considering_VVG the_AT data_NN stream_VV0 to_TO be_VBI fixed_VVN and_CC the_AT window_NN1 defined_VVN by_II its_APPGE two_MC end_NN1 points_NN2 ._. 
</s>
<s>
Such_DA ability_NN1 is_VBZ likely_JJ to_TO be_VBI governed_VVN by_II the_AT executive_NN1 system_NN1 ,_, which_DDQ in_II humans_NN2 start_VV0 to_TO develop_VVI at_II about_RG 6_MC years_NNT2 of_IO age_NN1 and_CC reach_VV0 full_JJ efficiency_NN1 at_II a_AT1 young_JJ adult_NN1 age_NN1 ._. 
</s>
<s>
SQL_NN1 is_VBZ an_AT1 extension_NN1 of_IO standard_JJ SQL_NN1 with_IW operators_NN2 for_IF stream_NN1 handling_NN1 ._. 
</s>
<s>
The_AT key_NN1 to_II the_AT model_NN1 is_VBZ the_AT ability_NN1 to_TO test_VVI these_DD2 statements_NN2 empirically_RR ._. 
</s>
<s>
To_TO accomplish_VVI this_DD1 and_CC also_RR evaluate_VVI the_AT performance_NN1 using_VVG a_AT1 ROC_NN1 curve_NN1 ,_, we_PPIS2 need_VV0 to_TO consider_VVI an_AT1 adaptive_JJ threshold_NN1 for_IF s(x)_NNU ._. 
</s>
<s>
The_AT Rapsodia_NP1 manual_NN1 &lsqb;_( 9_MC &rsqb;_) provides_VVZ details_NN2 and_CC examples_NN2 for_IF the_AT use_NN1 of_IO the_AT generator_NN1 ._. 
</s>
<s>
Therefore_RR ,_, if_CS a_AT1 reverse-neighbor_NN1 has_VHZ failed_VVN ,_, the_AT reverse-neighbor_JJ pointer_NN1 is_VBZ simply_RR deleted_VVN without_IW any_DD recovery_NN1 action_NN1 ._. 
</s>
<s>
First_MD ,_, the_AT analyst_NN1 must_VM determine_VVI the_AT degree_NN1 to_II which_DDQ she_PPHS1 can_VM trust_VVI that_CST all_DB of_IO the_AT relevant_JJ information_NN1 resides_VVZ in_RP at_RR21 least_RR22 one_MC1 of_IO the_AT connected_JJ repositories_NN2 ._. 
</s>
