期刊文献+

大语言模型赋能图书馆服务的测评及其应对策略

Evaluation and response strategies of empowering library services with the big language model
下载PDF
导出
摘要 大语言模型为图书馆智慧服务、精准服务的进步提供了巨大可能。文章通过测试国内外多个大语言模型在执行图书馆信息咨询服务时的全面性和准确性,为大语言模型赋能图书馆精准化智慧服务提供参考。以图书馆信息服务中的问题提示语为例,运用语言学方法,在词汇、句子及其关系的层次上分析、比较、统计这些输出结果的全面性和准确性及其原因。从词汇、句子及其关系的角度来看,大语言模型在图书馆精准服务方面的全面性和准确性程度来源于对其词汇、短语、句子、段落等实体及其关系标注的完善性以及算法的先进性与否。图书馆运用大语言模型赋能精准化、智慧化信息服务,建议参照多个大模型的输出结果以互补,再进一步配合相关文献资料予以分析、佐证,以力求所获知识的准确性和全面性,同时还应注重提示语的使用技巧。 The big language model provides great potential for the advancement of intelligent and precise services in libraries.The article tests the comprehensiveness and accuracy of multiple large language models at home and abroad in executing library information consulting services,providing reference for empowering libraries with precise and intelligent services through large language models.Taking the problem prompts in library information services as an example,using linguistic methods,analyze,compare,and statistically analyze the comprehensiveness,accuracy,and reasons of these output results at the level of vocabulary,sentences,and their relationships.From the perspective of vocabulary,sentences and their relationships,the comprehensiveness and accuracy of the big language model in the precise service of the library comes from the perfection of the annotation of its words,phrases,sentences,paragraphs and other entities and their relationships,as well as the progressiveness of the algorithm.The library utilizes a large language model to empower precise and intelligent information services.It is recommended to refer to the output results of multiple large models for complementarity,and further analyze and support them with relevant literature to strive for the accuracy and comprehensiveness of the knowledge obtained.At the same time,attention should also be paid to the use of language prompts.
作者 陈艳艳 CHEN Yanyan(Library,Henan University,Kaifeng 475004,China)
机构地区 河南大学图书馆
出处 《江苏科技信息》 2024年第12期78-82,96,共6页 Jiangsu Science and Technology Information
基金 河南省图书情报研究项目,项目名称:智慧服务情景下文献资源的精细化组织与精准化服务研究,项目编号:HNTQL-2023-008。
关键词 大语言模型 ChatGPT 图书馆赋能 信息服务 精准化服务 big language model ChatGPT library empowerment information service precision service
  • 相关文献

参考文献13

二级参考文献39

共引文献129

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部