基于内容相关性和时间分析的改进PageRank算法

摘要

Classical PageRank algorithm determines the PageRank values of web pages only according to the link structures of web pages, which usually results in theme-drift, that is the RageRank values of web pages that are not related to the theme content of these web pages are high.The improved PageRank algorithm distributes rank values according to the content similarity of linked web pages, and takes time factor of web pages into consideration, which can enhance the rank values of new relevant web pages.Experimental results show that the improved algorithm effectively solves the theme-drift problems of original algorithm, and make new pages rise in the ranking results.%传统PageRank算法仅仅通过网页链接结构来确定网页的PageRank值,这通常会导致主题漂移问题,即主题内容不相关的网页PageRank值较高.改进的PageRank算法根据链接的网页之间的内容相关性分配rank值,并且将网页的时间因素考虑在内,提高新的内容相关网页的rank值.实验结果证明,改进后的算法有效解决原算法的主题漂移问题,并使新网页在排序结果中上升.