路径聚类:在Web站点中的知识发现  

PATH CLUSTERING: DISCOVERING THE KNOWLEDGE IN THE WEB SITE

作　　者：王实[1] 高文[1] 李锦涛[1] 谢辉[1] 

机构地区：[1]中国科学院计算技术研究所,北京100080

出　　处：《计算机研究与发展》2001年第4期482-486,共5页Journal of Computer Research and Development

基　　金：国家"八六三"高技术研究发展计划基金资助!(86 3-30 6 -JD0 6 -0 3-4 )

摘　　要：用户对 Web站点的访问代表了用户对 Web站点上页面的访问兴趣 .这种兴趣程度可以通过用户对 Web站点上页面的浏览顺序表现出来 .在对 Web站点的访问日志进行事务识别后 ,可以根据群体用户对 Web站点的访问顺序进行聚类 ,即路径聚类 ,那么最终每一个聚类集就反映出该聚类集中的全体用户具有相似的访问兴趣 .为了得到这种根据用户访问兴趣而对用户集的划分 ,提出了 K- paths路径聚类方法 .在这种方法中 ,根据用户的访问兴趣定义了新的相似性测量手段和聚类中心 。When users access a Web site, the access of the users represents the interest of users in the Web pages of the Web site. Each user's interest can be manifested by the sequence of each user access. After processing the Log in the Web site and identifying each user access transaction, the access paths of all the users can be clustered. This is called path clustering. Each cluster can then represent the similar access interest of the users in the cluster. Presented in this paper is a new clustering approach: K paths to partition the users' access according to the interest of the users. In this approach, according to the requirement of the clustering, the new method is defined to measure similarity and to get the center of a cluster. The experiment shows that this approach is successful.

关 键 词：数据挖掘 WEB站点 知识发现 路径聚类 WWW INTERNET 

分 类 号：TP393.4[自动化与计算机技术—计算机应用技术]