一种基于网页质量的PageRank算法改进

随着Web技术的发展和Web上越来越多的各种信息,如何提供高质量、相关的查询结果成为当前Web搜索引擎的一个巨大挑战.PageRank和HITS是两个最重要的基于链接的排序算法并在商业搜索引擎中使用.然而,在PageRank算法中,每个网页的PR值被平均地分配到它所指向的所有网页,网页之间的质量差异被完全忽略.这样的算法很容易被当前的Web SPAM攻击.基于这样的认识,提出了一个关于PageRank算法的改进,称为Page Quality Based PageRank(QPR)算法.QPR算法动态地评估每个网页的质量,并根据网页的质量对每个网页的PR值做相应公平的分配.在多个不同特性的数据集上进行了全面的实验,实验结果显示,提出的QPR算法能大大提高查询结果的排序,并能有效减轻SPAM网页对查询结果的影响.

An Improvement of Page Rank Algorithm Based on Page Quality

With the development of Web and more information distributed on the Web,how to provide best quality search results is a big challenge for current search engines.PageRank and HITS are the two most important link-based ranking algorithms and have been used in some commercial search engines.However,in the PageRank algorithm,PageRank value of each page is distributed to all its citation pages in average and thus the quality difference between pages is totally ignored during the distribution process.Such algorithm easily becomes the target of Web spam.In this paper,an improvement of PageRank algorithm is proposed based on page quality(called QPR algorithm).The quality of each page is evaluated dynamically and its PR value is distributed to all its citation pages based on page quality accordingly.Experiments are conducted comprehensively on different datasets and experimental results demonstrate that the proposed ranking algorithm could improve final search results ranking significantly and also alleviate the influence of spam pages effectively.