《大语言模型的外语教学与研究应用》(又称《棋盘书》)
点击阅读2024年两次“我要读文献”在线交流文字实录。
点击下图进入PromptBank提示语库
常用大语言模型
国内
国际
PromptBank for Language Teachers and Researchers(语言教学与研究提示语库)
PromptBank: A repository of large language model prompts for language research and learning purposes
北外语料库团队特设立本网页,旨在收集、整理和分享用于语言学研究和语言教学的大语言模型(如ChatGPT、Claude、Bard、Midjourney、文心一言、讯飞星火、通义千问等)提示语,以推动语言学界对大语言模型的应用。请各位同仁帮忙共同测试。如有好的提示语愿意分享,或有任何问题可联系我们bfsucrg@sina.com。The Corpus Research Group at Beijing Foreign Studies University (BFSU) has established this webpage with the aim of collecting and sharing prompts for large language models (such as ChatGPT, Claude, Bard, Midjourney, Wenxin Yiyan/ERNIE Bot, iFlyTek SparkDesk, Tongyi Qianwen/Alibaba chatbot, among others) that are composed for linguistic research and language teaching. Our goal is to promote the use of large language models in (applied) linguistics. Should you have any valuable prompts to share or questions, please feel free to email us at bfsucrg@sina.com.
This webpage is designed to enhance linguistic studies through the power of AI. The notion of "AI for Linguistics" has become more accessible and practical than ever, particularly with the advent of user-friendly generative AI readily available to all.
#1: 事件语义学分析(Event semantics analysis)
View the prompts and responses from ChatGPT regarding the event semantics of assessment markers in English.
(编者:孔蕾教授;上传时间:2023年8月12日;测试平台:GPT3.5/GPT4)
#2:英汉平行文本句对齐操作
A01A.EN.txt contains the original text, and A01A.ZH.txt is its corresponding Chinese translation. Could you align the two texts at the sentence level, and then create a new line for each aligned sentence pair?
(编者:许家金;上传时间:2023年8月10日;测试平台:Claude2)
#3:教材文化呈现研究
Can you mark up the segments representing the five cultural dimensions (i.e., cultural products, cultural practices, cultural perspectives, cultural communities, and cultural persons) in the following text? [Please paste your text here.]
(编者:许家金;上传时间:2023年8月10日;测试平台:GPT3.5/GPT4)
#4: Can you classify the following items into the five categories of "self-esteem, love and belonging, physiological needs, personal and financial safety, and self-actualization"?
lose weight, stop procrastinating, write a book, get a tattoo, travel the world, see the northern lights, learn Spanish, learn to play the guitar, get a job, run a marathon, learn French, skydive, be more confident, write a novel, learn Japanese, get in shape, quit smoking, start my own business, learn to cook, fall in love, be happy, get married, kiss in the rain, make new friends, drink more water, exercise regularly, save money, buy a house, get out of debt
(编者:许家金;上传时间:2023年8月11日;测试平台:GPT3.5/GPT4)
#5 Can you annotate the conceptual metaphor cases in the following text?[您可在此处粘贴一段文字作文待分析材料]
(编者:许家金;上传时间:2023年8月12日;测试平台:GPT4)
#6:Could you please generate a Correspondence Analysis plot using the provided data?
(编者:许家金;上传时间:2023年8月13日;测试平台:GPT4,启用Code Interpreter)
#7: According to Google Scholar, which are the 10 most downloaded linguistics research articles in 2023?
(编者:许家金;上传时间:2023年8月13日;测试平台:Gemini/Bard)
#8: Could you apply the Initiation-Response-Feedback (IRF) framework to analyze the conversation provided and annotate the text in XML format?
老师:涵涵,这头上这里是怎么啦?
涵涵:我在家不小心撞到了。
老师:哦,磕到了,是吧?
涵涵:我昨晚为了去关灯,不小心撞到晾衣架上了。不过,没事的,你不碰它,它自己就会好的。
(编者:许家金;上传时间:2023年8月13日;测试平台:GPT3.5/GPT4)
#9: Could you analyze the given English text for various linguistic features, specifically: Word Count, Sentence Count, Verb Phrase Count, Clause Count, T-unit Count, Dependent Clause Count, Complex T-unit Count, Coordinate Phrase Count, and Complex Nominal Count? Please present the results in a tabular format. [Please paste your text here.]
[NB: Results should be verified using standard syntactic complexity analysis methods.]
(编者:许家金;上传时间:2023年8月15日;测试平台:GPT3.5)
#10: Could you analyze the given English text for the following 14 syntactic complexity indices: Mean length of sentence (MLS), Mean length of T-unit (MLT), Mean length of clause (MLC), Clause per sentence (C/S), Verb phrase per T-unit (V/T), Clause per T-unit (C/T), Dependent clause per clause (DC/C), Dependent clause per T-unit (DC/T), T-unit per sentence (T/S), Complex T-unit ratio (CT/T), Coordinate phrase per T-unit (CP/T), Coordinate phrase per clause (CP/C), Complex nominal per T-unit (CN/T), and Complex nominal per clause (CN/C)? Please present the results in a tabular format. [Please paste your text here.]
[NB: Results should be verified using standard syntactic complexity analysis methods.]
(编者:许家金;上传时间:2023年8月15日;测试平台:GPT3.5)
#11: Can you list all the inflectional forms of the following verbs in the form of a table, the first column is the base form, all the other columns are inflectional forms?
SHOW, SEE, FIND, SUGGEST, INDICATE, CONSIDER, REPORT, DESCRIBE, OBSERVE, NOTE, ASSUME, DEMONSTRATE, KNOW, EXPECT, PREDICT, ESTIMATE, EXPLAIN, REVEAL, PROPOSE, and MEAN
(编者:许家金;上传时间:2023年8月15日;测试平台:GPT3.5/GPT4; Gemini/Bard; Claude2;通义千问)
#12: Can you create a tree diagram and perform a transformational-generative syntactic analysis of the sentence "colorless green ideas sleep furiously"?
(编者:许家金;上传时间:2023年8月16日;测试平台:GPT3.5/GPT4; Gemini/Bard; Claude2)
#13: Can you compute the mean dependency distance (MDD) of the provided text? [Please paste your text here.]
(编者:许家金;上传时间:2023年8月16日;测试平台:Gemini/Bard)
#14: a man picking pears on a ladder in a tree.
[For eliciting spoken narrative data in an L2 picture description task. See the L2 Pear Story Project.]
(编者:许家金;上传时间:2023年8月17日;测试平台:通义万相)
#15: Could you annotate the moves of the provided abstract and represent the annotation in XML format? [Please paste your text here.]
(编者:许家金;上传时间:2023年8月18日;测试平台:GPT3.5/GPT4; Gemini/Bard; Claude2;文心一言)
#16: Can you summarize the theme of the provided text in a three-word phrase?
A very important thing is to equip parents with necessary techniques of conducting assessment ... When this [getting parents involved in assessment] is utterly necessary, say when I'm unable to conduct assessment online, I'll make a video to train parents to assess their children on a particular language task. For example, in a role-playing activity, what is outstanding performance like? What is the average like? What are commonly anticipated problems? ... I invite parents to join us in short extracurricular sessions where I conduct assessment on one student's performance as a demo, and then encourage a parent to have a go ... A boy in my class told me that he thought his mother was more capable of "being a teacher at home" after the "training".
ChatGPT: Empowering Parental Assessment
(编者:许家金;上传时间:2023年8月18日;测试平台:GPT4)
#17: Could you annotate the provided text with its parts of speech using the TreeTagger tagset?
A very important thing is to equip parents with necessary techniques of conducting assessment ... When this [getting parents involved in assessment] is utterly necessary, say when I'm unable to conduct assessment online, I'll make a video to train parents to assess their children on a particular language task.
Claude2: Here is the text with part-of-speech tags using the TreeTagger tagset:
A/DT very/RB important/JJ thing/NN is/VBZ to/TO equip/VB parents/NNS with/IN necessary/JJ techniques/NNS of/IN conducting/VBG assessment/NN ... When/WRB this/DT [getting/VBG parents/NNS involved/VBN in/IN assessment/NN ]/NP is/VBZ utterly/RB necessary/JJ ,/, say/VB when/WRB I'm/PRP unable/JJ to/TO conduct/VB assessment/NN online/RB ,/, I'll/MD make/VB a/DT video/NN to/TO train/VB parents/NNS to/TO assess/VB their/PRP$ children/NNS on/IN a/DT particular/JJ language/NN task/NN ./.
(编者:许家金;上传时间:2023年8月18日;测试平台:GPT3.5; Claude2)
#18: Both of the uploaded texts utilize a third-person omniscient point of view. However, could you please provide me with more information about the texts "Yangs001.txt" and "Hawkes 001.txt" so that I can discern the nuanced differences between their points of view?
Claude2 response:
"Hawkes001.txt" uses a more formal, academic tone. The narrator provides context about the background and origins of the story, referring to literary elements like the "allegory of the jade" and "recourse to names." There are also explanatory footnotes citing sources. Overall, the narrator seems more detached and analytical.
"Yangs001.txt" has a more conversational, casual tone. The narrator directly addresses the reader as "Worthy Readers" and "Gentle Reader," inviting them into the storytelling. There is more commentary woven into the narration, such as "Strange to relate" and predictions about what's to come ("I imagine this story should have more fine points than the usual run of breeze-and-moonlight tales"). The narrator seems more informal and involved.
In summary: "Hawkes001.txt" has a more formal, academic narrator who analyzes the literary aspects of the story. "Yangs001.txt" has a more casual, conversational narrator who draws the reader into the storytelling. The main narrative content is similar between the two, but the narrator's voice differs in tone and approach.
(编者:许家金;上传时间:2023年8月19日;测试平台:Claude2)
#19:请以五点式李克特量表设计一份包含30个问题的问卷。问卷的主题为“大学生对使用ChatGPT辅助外语教学的态度”。问卷设计完成后,请提供问卷信度效度检验的分析方法。
Please create a 30-question survey using a five-point Likert scale, centered around the theme: 'University students' perceptions of using ChatGPT in aiding foreign language learning'. Once the survey is formulated, please suggest methods to evaluate its reliability and validity.
(编者:许家金;上传时间:2023年8月11日;测试平台:GPT4;文心一言;讯飞星火)
#20a: Could you compare the two texts and provide a comprehensive list of differences between the two versions in the order they appear in the original texts, using a table format?
#20b: Could you compare the two texts and provide a comprehensive list of differences between the two versions and categorize them stylistically and provide examples of change for each category?
(编者:许家金;上传时间:2023年8月19日;测试平台:POE-Assistant)
#21: Prototype bird experiment. Please rank each of the following four items on a scale of 1 to 7 as a good example of bird.
A) 1 2 3 4 5 6 7
|
B) 1 2 3 4 5 6 7
|
C) 1 2 3 4 5 6 7
|
D) 1 2 3 4 5 6 7
|
The images were generated by Alibaba Tongyi Wanxiang.
#22: Reading comprehension task. Merchant and passenger ships are generally required to have a life preserver for every person aboard and in many cases, a certain percentage of smaller sizes for children. According to United States requirements, life preservers must design, reversible capable of being quickly adjusted to fit the uninitiated individual, and must be so designed as to support the wearer in the water in an upright or slightly backward position.
1. What is the passage mainly about?
(编者:许家金;上传时间:2023年8月20日;测试平台:通义万相)The images were generated by Alibaba Tongyi Wanxiang.
#23: I am a dictionary entry writer, and want to write an English-Chinese bilingual entry for the word "agricultural" for Chinese vocational students specializing in agriculture and forestry. Below is a sample entry for the word "plant". Please adhere to the format of the entry "plant" closely."[Please right click to download the full prompt.] [Please right click to download the GPT generated entry for "agricultural" here.]
(编者:许家金;上传时间:2023年8月21日;测试平台:GPT3.5/通义万相)
|
|
#23 dictionary entry illustration image generated by Alibaba Tongyi Wanxiang using the prompt 'agricultural.' |
#24 A visual representation of the first point summarized by Claude2, i.e.Tongyi Wanxiang prompt: China's continued economic growth and development. GDP growth reached 6.6%, new jobs were created, consumption expanded. It discusses efforts to boost innovation, entrepreneurship, reduce poverty, improve living standards. |
#24: How does the given text portray China?
[2019 Government Work Report, English translation]; Claude2 response: [Right click to download the text.]
(编者:许家金;上传时间:2023年8月21日;测试平台:Claude2/通义万相)
#25: What image comes to mind when you read the opening paragraphs of "Alice's Adventures in Wonderland"?
Alice was beginning to get very tired of sitting by her sister on the bank, and of having nothing to do: once or twice she had peeped into the book her sister was reading, but it had no pictures or conversations in it, 'and what is the use of a book,' thought Alice 'without pictures or conversation?'
So she was considering in her own mind (as well as she could, for the hot day made her feel very sleepy and stupid), whether the pleasure of making a daisy-chain would be worth the trouble.
[This can be used as a warm-up or lead-in activity in an English class.]
(编者:许家金;上传时间:2023年8月22日;测试平台:通义万相)
#26: Could you transform the text into a well-structured table using tab-delimited format? + Speech Writing Total number of adverbials 712 753 Realization types: Prepositional phrases 258 336 Closed-class items 278 246 Open-class adverbs and adverb phrases 76 66 Finite clauses 70 60 Nonfinite and verbless clauses 12 33 Noun phrases 18 12
[The source of the dataframe is provided in the image.]
(编者:许家金;上传时间:2023年8月22日;测试平台:GPT4/Claude2)
#27: Do "room service" and "客房服务" have the same meaning?
|
|
room service |
客房服务 |
(编者:许家金;上传时间:2023年8月22日;测试平台:通义万相)
#28a: Please format the references using MLA style?
Simpson, R. & J. Swales (eds.). 2001. Corpus Linguistics in North America [C]. Ann Arbor: University of Michigan Press.
Sinclair, J. 1991. Corpus, Concordance, Collocation [M]. Oxford: Oxford University Press.
Thorndike, E. 1921. The Teacher's Word Book [M]. New York City: Teachers College, Columbia University.
West, M. 1931. Notes, news and clippings [J]. The Modern Language Journal 15(8): 638-647.
#28b: Please format the references using APA style?
Simpson, R., and J. Swales, editors. Corpus Linguistics in North America. University of Michigan Press, 2001.
Sinclair, J. Corpus, Concordance, Collocation. Oxford University Press, 1991.
Thorndike, E. The Teacher's Word Book. Teachers College, Columbia University, 1921.
West, M. "Notes, news and clippings." The Modern Language Journal, vol. 15, no. 8, 1931, pp. 638-647.
(编者:许家金;上传时间:2023年8月23日;测试平台:GPT4/Claude2)
#29: Could you provide 40 commonly used academic English sentence structures that can be used to critique limitations of previous research?
(编者:许家金;上传时间:2023年8月23日;测试平台:GPT3.5、GPT4、Claude2、Gemini/Bard、文心一言、讯飞星火、通义千问)
#30: Could you provide a cheat sheet of the sub-fields within Applied Linguistics?
(编者:许家金;上传时间:2023年8月23日;测试平台:GPT4、Claude2、Gemini/Bard)
#31: Can you simplify this scientific report into a 250-word English summary suitable for Chinese freshmen?
[Link to the report: https://www.nature.com/articles/s41598-022-20841-0]
(编者:许家金;上传时间:2023年8月23日;测试平台:Claude2)
#32: Could you conduct a sentiment analysis on the two translations (i.e., Yangs001.txt and Hawkes001.txt) of the same source text, compare their sentiment differences, and provide examples to support your findings?
[Please view the LLM response here.]
|
|
The opening scene of Yangs' Ch1 of Hong Lou Meng visualized using AI image. |
The opening scene of Hawkes' Ch1 of Hong Lou Meng visualized using AI image. |
(编者:许家金;上传时间:2023年8月24日;测试平台:Claude2)
#33: Could you create a multiple-choice exercise with five questions based on the provided text, tailored for second-year Chinese college students studying English? After the exercise, please provide the answers and explain why the other three choices are incorrect. [Please paste your text here.]
(编者:许家金;上传时间:2023年8月24日;测试平台:GPT3.5/GPT4; Gemini/Bard; Claude2)
#34: Role-playing Chatbots. E.g. You have the opportunity to engage in conversation with Elon Musk, Tim Cook, Bill Gates, Lady Gaga, William Shakespeare, and Albert Einstein, as shown in the screenshot.
(编者:许家金;上传时间:2023年8月24日;测试平台:https://www.character.ai)
#35: Could you compare the three texts in terms of the following 14 syntactic complexity indices: Mean length of sentence (MLS), Mean length of T-unit (MLT), Mean length of clause (MLC), Clause per sentence (C/S), Verb phrase per T-unit (V/T), Clause per T-unit (C/T), Dependent clause per clause (DC/C), Dependent clause per T-unit (DC/T), T-unit per sentence (T/S), Complex T-unit ratio (CT/T), Coordinate phrase per T-unit (CP/T), Coordinate phrase per clause (CP/C), Complex nominal per T-unit (CN/T), and Complex nominal per clause (CN/C)? Please present the result in a tabular format.
[Please view the result here. Results should be verified using standard syntactic complexity analysis methods.]
(编者:许家金;上传时间:2023年8月25日;测试平台:Claude2)
#36: Could you generate a topic model for the provided text using the LDA algorithm and display the results in a table and create a word cloud image using the topic words as well?
(编者:许家金;上传时间:2023年8月25日;测试平台:GPT4,启用Code Interpreter)
#37: Could you analyze the overall research trend based on the bibliographic data?
(编者:许家金;上传时间:2023年8月28日;测试平台:GPT4,启用Code Interpreter)
#38: Could you perform a sentiment analysis on the provided text and plot a heatmap based on the sentiment analysis results?
(编者:许家金;上传时间:2023年8月28日;测试平台:GPT4,启用Code Interpreter)
#39: Could you perform a correlation analysis between the "score" and the "connectives" based on the provided data?
(编者:许家金;上传时间:2023年8月28日;测试平台:GPT4,启用Code Interpreter)
#40: Could you visualize this relationship between the "score" and the "connectives" based on the provided data using a scatter plot?
(编者:许家金;上传时间:2023年8月28日;测试平台:GPT4,启用Code Interpreter)
#41: 请基于以下信息编写APA格式文献条目
Register variation in written contact varieties of English: a multidimensional analysis
Haidee Kruger, Bertus Van Rooy
Original languageEnglish
Pages (from-to)214-242
Number of pages29
JournalEnglish World-Wide
Volume39
Issue number2
Publication statusPublished - 2018
结果:Kruger, H., & Van Rooy, B. (2018). Register variation in written contact varieties of English: a multidimensional analysis. English World-Wide, 39(2), 214-242.
(编者:许家金;上传时间:2023年9月24日;测试平台:GPT4)
相关资源
《语料库研究方法》配套网页
《语料库与话语研究》配套网页
《语料库与中国学习者英语口语研究》配套网页