摘要
文章通过分析高校图书馆个性化推荐系统的应用现状和存在问题,设计了融合用户需求特性和行为习惯的推荐系统架构。该架构采用B/S架构,通过深度分析文献摘要内容,实现与用户兴趣的精准匹配,包括数据采集、预处理、摘要生成、内容过滤、用户兴趣建模、个性化推荐和反馈调整等核心模块。实验结果表明,该系统能显著提升文献推荐的准确性,MAE指数提高9.8%,时间复杂度降低19.7%,有效提高了推荐效率,为高校图书馆提供个性化推荐解决方案,具有重要的实践价值和理论意义。
By analyzing the current application status and existing problems of personalized recommendation systems in university libraries,this paper designs a recommendation system architecture that integrates user demand characteristics and behavioral habits.The architecture employs a B/S(Browser/Server)structure,achieving precise matching with user interests through in-depth analysis of document abstract content.It encompasses core modules such as data collection,preprocessing,abstract generation,content filtering,user interest modeling,personalized recommendation,and feedback adjustment.The results demonstrate that the system can significantly enhance the accuracy of document recommendations,with an increase of 9.8%in the MAE(Mean Absolute Error)index and a reduction of 19.7%in time complexity,effectively improving recommendation efficiency.This provides a personalized recommendation solution for university libraries and holds significant practical value and theoretical significance.
出处
《河北科技图苑》
2025年第4期84-90,48,共8页
Hebei Journal of Library and Information Science
基金
广西民族师范学院人才项目“民族地方高校图书馆的阅读服务现状及创新服务对策研究”(2023SBNGCC014)
广西民族师范学院2022年度校级科研项目“广西民族师范学院民族文献资源现状分析与共建共享保障机制研究”(2022YB038)的研究成果
关键词
内容过滤
图书馆
个性化推荐
文献资源
Content Filtering
Library
Personalized Recommendation
Document Resources
