摘要
[目的/意义]评估中文学术文献阅读的AI工具效果,为工具改进提供参考。[方法/过程]选取CNKI AI学术研究助手与文心一言智能体阅读助手Plus,从功能性、效率性、易用性及稳定性与安全性四个维度进行测评和比较。[结果/结论]两款工具各有优势,通过测评发现,CNKI助手在复杂检索方面的准确率达到90%以上,而文心助手在处理多篇文献时效率提升约40%。研究揭示了不同工具的特点,为AI在学术文献阅读领域的优化提供了指导。通过对比分析,有助于开发者针对用户需求进行更精准的改进,提升学术研究效率。提出了关于学术研究助手未来研发的具体建议:加强文献管理与检索功能的整合,提升文献信息抽取与分析的全面性,优化用户体验与交互设计,保障系统稳定性与数据安全性,推动AI工具与学术研究的深度融合。
[Purpose/significance]To evaluate the effectiveness of AI tools for reading Chinese academic literature and provide references for tool improvement.[Method/process]Two representative AI tools,CNKI AI Academic Research Assistant and ERNIE Bot Plus,were selected for evaluation across four dimensions:functionality,efficiency,ease of use,and stability&security.[Result/conclusion]Both tools exhibit their respective advantages.The CNKI Assistant performs better in literature management,complex retrieval,and information annotation,while the ERNIE Bot Plus excels in processing speed and multi-document analysis.This study unveils the characteristics of different tools,offering guidance for the optimization of AI in the field of academic literature reading.Through comparative analysis,it assists developers in making more precise improvements tailored to user needs,thereby enhancing the efficiency of academic research.Suggestions for the future development of academic research assistants have been proposed:strengthening the integration of literature management and retrieval functions,enhancing the comprehensiveness of literature information extraction and analysis,optimizing user experience and interaction design,ensuring system stability and data security,and promoting the deep integration of AI tools and academic research.
作者
刘奇扉
Liu Qifei(Library of Wuhan University of Science and Technology,Wuhan 430081,China)
出处
《图书情报研究》
2025年第3期32-39,共8页
Library and Information Studies
作者简介
刘奇扉,女,1978年生,副研究馆员,研究方向为数字图书馆、图书馆服务创新,发表论文数十篇。
