融合反馈信息与内容相关度的PageRank改进算法

Improved PageRank algorithm of merging feedback information and topical relationship

为了提高网页排序算法的效率,提高搜索引擎的检索质量,提出了融合反馈信息与内容相关度的PageRank改进算法。利用向量空间模型计算网页之间的主题相关性,得到网页的主题相关度权值。通过对网页被点击次数进行统计分析,得到网页点击量的增量权值。将这两个权值结合共同影响网页的PR(PageRank)值分配。通过仿真实验得到运用该算法后的实验结果,与其它算法的实验结果进行了比较,验证了该算法优于其它算法。

To improve the efficiency of web sorting algorithm and improve the retrieval quality of the search engine,an improved Page-Rank algorithm merging feedback information and topical relationship is presented.Firstly,the theme relatedness weights are gotten by calculating the topical relationship between web pages using the vector space model.Then,through the statistical analysis of clicks,the clicks incremental weights are obtained.Finally,these two weights are merged to influence the distribution of the PR value.Comparing the experimental results of simulation,the advantage of this algorithm is proved.