摘要
根据推荐系统对用户(商品)聚类的要求,探讨采用用户(网页)兴趣度进行聚类分析的合理思想。通过用户浏览时间、浏览行为以及网页信息量差异等因素的对比,得出用户对某类商品的兴趣度计算方法。借助阈值的设定,定义了用户感兴趣的商品集、商品的感兴趣用户集和兴趣相似的用户集,得到了基于用户兴趣度的用户聚类的一般过程,具有一定的推广价值和借鉴意义。
Based on the need of user or merchandise clustering in recommender systems,the idea of user interest level or Web interest level is proposed in the paper.Through comparing users’ browsing time,browsing action and the amount of information,a method of computing user interest level to a class of merchandise is put forward.By setting thresholds,the interested merchandise set of users,interested user set of merchandise and user set of similar interest are defined.So the general steps of users clustering based on user interest level are gotten,which has a meaning of promotion and reference.
出处
《计算机工程与应用》
CSCD
北大核心
2011年第7期226-228,共3页
Computer Engineering and Applications
关键词
推荐系统
用户兴趣度
聚类分析
recommender systems
user interest level
clustering analysis
作者简介
崔春生(1974-),男,博士,主要研究方向:决策理论与方法、电子商务智能推荐系统
吴祈宗(1947-),男,教授,博士生导师。