以ChatGPT为代表的自然语言AI在数据库内容检索与生成中的应用——以全国报刊索引为例  

The Application of Natural Language AI Represented by ChatGPT in Database Content Retrieval and Generation:Taking National Newspaper Index as an Example

作　　者：戴梦菲 Dai Mengfei(Shanghai Library,Shanghai 200031)

机构地区：[1]上海图书馆,上海200031

出　　处：《情报探索》2024年第5期103-108,共6页Information Research

摘　　要：[目的/意义]旨在探讨以ChatGPT为代表的自然语言模型在图书馆馆藏资源数据库中的应用,为人工智能时代图书馆数据库的发展方向提供借鉴。[方法/过程]采用案例分析法,以上海图书馆的全国报刊索引数据库为研究对象,分析ChatGPT的基本工作原理,并从自然语言问答、信息整合、大纲生成等方面,评估人工智能对于传统的数据库内容检索与生成方式的优势,及可能产生的问题。[结果/结论]该技术在处理自然语言查询方面表现出色,能够有效提高数据库的检索效率,并为读者提供创造性的内容生成。但其应用也面临技术整合、内容真实性与专业性等方面的问题和挑战,需要进一步研究并制定规范。[Purpose/significance]This paper aims to explore the application of natural language model represented by ChatGPT in library collection resource database and provide references for the development direction of library database in the era of artificial intelligence.[Method/process]Using the case analysis method and taking the database of National Newspaper Index in Shanghai Library as the research object,this paper analyses the basic working principle of ChatGPT and evaluates the advantages and possible problems of artificial intelligence on traditional database content retrieval and generation methods from the aspects of natural language question-and-answer,information integration,outline generation.[Result/conclusion]The technology is excellent in processing natural language queries,and it can effectively improve the retrieval efficiency of the database and provide readers with creative content generation.However,its application also faces the problems and challenges in aspects of technology integration,content authenticity and professionalism,which needs to be further studied and formulated.

关 键 词：图书馆服务 人工智能 生成式语言模型 数据库检索 

分 类 号：G250.7[文化科学—图书馆学]