信息素养教育领域中GPT-4生成论文引言与学者撰写引言对比研究

A Comparative Study of GPT-4 Generated Introduction and Scholars’Written Introduction in the Field of Information Literacy Education

[目的/意义]旨在对比分析学者撰写的图书馆信息素养教育论文引言与GPT-4生成引言的异同,并探讨其对图书馆AI素养教育的启示。[方法/过程]对比96篇高被引论文的学者撰写引言与GPT-4生成的引言,利用机器学习分类方法和内容分析技术,深入探讨两者在语言表达、逻辑结构、信息准确性等方面的异同。[结果/结论]学者撰写的引言与GPT-4生成的引言在语言表达上具有一定的相似性,但在逻辑结构、信息准确性等方面存在差异。GPT-4更倾向于使用特定词汇和短句,内容泛化且事实性信息生成能力一般。图书馆应从普及AI知识、开设AI课程、组建教学团队等方面提升馆员和读者人工智能素养。

[Purpose/significance]This paper aims to compare and analyze the similarities and differences between the introduction of the paper on library information literacy education written by scholars and the introduction of GPT-4 generation,and explore its enlightenment to library AI literacy education.[Method/process]By comparing the introduction written by the scholars in 96 highly cited papers and the introduction generated by GPT-4,the paper uses machine learning classification method and content analysis technology to deeply explore the similarities and differences in language expression,logical structure and information accuracy.[Result/conclusion]The introduction written by scholars and the introduction generated by GPT-4 have some similarities in language expression,but there are differences in logical structure and information accuracy.GPT-4 tends to use specific words and short sentences,generation ability of content generalization and factual information is general.Libraries should promote the artificial intelligence literacy of librarians and readers from the aspects of popularizing AI knowledge,setting up AI courses,and building teaching teams.