用户协同过滤个性化图书推荐算法的改进与实现  

Improvement and Implementation on Personalized Book Recommendation Algorithm Based on User Collaborative Filtering

作　　者：刘军军[1] LIU Junjun

机构地区：[1]渤海大学图书馆,辽宁锦州121013

出　　处：《图书情报导刊》2021年第1期38-42,共5页Journal of Library and Information Science

摘　　要：读者想要迅速从海量图书资源中找到需要的图书是其面临的难题,个性化图书推荐服务为解决这一难题提供了有效手段。针对传统用户协同过滤算法存在的问题,依据用户协同过滤推荐算法基本原理,以基本数学模型为基础,从解决数据稀疏性、用户兴趣随着时间迁移而衰减、用户特性影响用户兴趣等方面进行了改进,并设计了算法实现流程,为应用系统开发提供解决方案。研究结果表明,通过对用户协同过滤个性化图书推荐算法的改进,有效提高了图书资源检索和推荐的智能水平,满足了用户的个性化需求,解决了软件开发关键性技术问题。It is difficult for readers to quickly find the books they need from the massive book resources.Personalized book recommendation service provides an effective means to solve this problem.Aiming at the problems existing in traditional user collaborative filtering algorithm,according to the basic principle of user collaborative filtering recommendation algorithm and based on the basic mathematical model,the improvement is made from the aspects of solving data sparsity,user interest decaying with time migration,user characteristics affecting user interest,etc.,and the algorithm implementation process is designed to provide solutions for application system development.The research results show that the intelligent level of book resource retrieval and recommendation is effectively improved through the improvement of personalized book recommendation algorithm through user collaborative filtering,which meets the personalized needs of users and solves the key technical problems of software development.

关 键 词：用户协同过滤 个性化图书推荐 算法改进 实现流程 

分 类 号：G252.62[文化科学—图书馆学]