基于多标签的电子图书资源协同过滤推荐算法研究  

Research on Collaborative Filtering Recommendation Algorithm of E-book Resources Based on Multiple Tags

作　　者：吴晓倩[1] 陈诚[2] 刘小侠 WU Xiaoqian;CHEN Cheng;LIU Xiaoxia(School of Public Health and Health Management,Anhui Medical College,Hefei 230601,China;Library of Anhui Medical College,Hefei 230601,China)

机构地区：[1]安徽医学高等专科学校公共卫生与健康管理学院,安徽合肥230601 [2]安徽医学高等专科学校图书馆,安徽合肥230601

出　　处：《佳木斯大学学报（自然科学版）》2024年第11期22-25,共4页Journal of Jiamusi University：Natural Science Edition

摘　　要：针对基于标签的电子图书推荐算法中存在的冷启动、稀释性、时间权重和隐私保护问题,设计了一种综合多标签的电子图书资源协同过滤推荐算法。该算法采用多维度相似度计算,结合用户自身标签和资源标签的相似度,提高推荐准确性,更精准地为用户推荐适合的电子图书。同时,引入差分隐私保护机制,有效保护用户隐私。实验结果表明,该算法在准确率、召回率和F1值上表现优异,为图书馆资源推荐提供了一种高效且隐私保护的解决方案。To address the issues of cold start,dilution,time-weighted relevance,and privacy protection in tag-based e-book recommendation algorithms,we design a comprehensive multi-label collaborative filtering recommendation algorithm for e-book resources.This algorithm employs multi-dimensional similarity calculations,integrating the similarity between users'own tags and resource tags,to enhance the accuracy of recommendations and provide users with more precise e-books that suit their interests.Additionally,a differential privacy protection mechanism is introduced to effectively safeguard user privacy.Experimental results demonstrate the algorithm's outstanding performance in terms of precision,recall,and F1 score,offering an efficient and privacy-preserving solution for library resource recommendations.

关 键 词：标签 协同过滤 推荐算法 差分隐私 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]