摘要
提出一种新的用户访问聚类方法,采用模糊聚类原则,将已有的用户访问主页的URL记录快速、有效地为有相似兴趣、爱好的访问者访问记录群聚类 利用这种方法对访问用户聚类,再根据其聚类结果推荐给用户相应的动态链接.
This paper presents a novel clustering method that takes web sever logs as input and outputs the users partitions .First we define the hierarchical similarity of URLs to distinguish the users log session .Next by using the clustering method in Fuzzy logic we can find the clustering nucleus efficiently. Having finished the clustering procedure ,this method can provide the recommendable links for surfing users.
出处
《西北民族大学学报（自然科学版）》
2004年第2期57-60,89,共5页
Journal of Northwest Minzu University(Natural Science)
关键词
模糊聚类
智能链接
数据挖掘
fuzzy cluster
intellectual link
data mining
