一种支持个性化Web推荐的聚类分析  

A Cluster Analysis for Personalized web Recommendation

作　　者：陈雪刚[1,2] 杨磊[1] 

机构地区：[1]湖南大学计算机与通信学院,湖南长沙410082 [2]湘南学院计算机系,湖南郴州423000

出　　处：《计算技术与自动化》2010年第1期130-133,共4页Computing Technology and Automation

摘　　要：提出基于关联的聚类分析方法,挖掘具有相似访问兴趣的用户访问模式,分离不相关的用户模式,并提出基于关联的聚类算法。实验证明,该算法大量减少不相关的用户访问模式,提高个性化推荐质量。为进一步研究个性化推荐技术奠定基础。In order to provide better personalized recommendation service, and cluster analysis methods based on association has been improved, which mining model of the users access with interests of similar access, and user model of non--related is separated a clustering. An algorithm based on association is proposed. It is proved that the algorithm can reduce a large number of user model of non--related, so improve the quality of the personalized recommendations. It is laid the foun- dation to further study the personalized recommendation technology.

关 键 词：关联规则 数据挖掘 聚类 个性化推荐 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]