Clustering with Weighted Hyperlink and Sub Similarity Matrix

Now, effectively searching web page becomesmore and more critical as the online information con-tinues to overload the web. Asolution is to make areasonable web page clustering analysis . But there isanother problem,because we must find out the webstructure and page hyperlinks , especially si milarityamongthe pages . At the same ti me ,there are someother very i mportant information such as the i mpor-tance of each page ,page content and so on[1 -3].Hyperlink analysis has already been applied …     

PageCluster:一种Web页面层次聚类方法

提出了Web页面聚类算法PageCluster及相应的改进算法ImPageCluster。该方法在兼顾Web站点结构和页面链接的同时,基于各个页面的重要程度对各个超链接进行赋权。与传统聚类算法相比,该算法不需要事先给定相似度阈值。实验结果证实了该算法的可行性和高效性。