利用蚁群算法对PageRank算法的改进

Improvement of PageRank algorithm by ant colony algorithm

在PageRank算法的基础上应用蚁群算法的主要思想,对网页按关键字检索后被点击的次数进行统计,根据其在初始排序结果中的位置对网页进行分类,通过给定的函数变换对按照以上两个因素统计分析所得的结果进行运算,将其作为网页与关键字关联度的一个评判依据。从而对网页的权值(PR值)进行迭代修正,并返回一个新的排序结果。通过模拟实验表明,此方法在使得返回结果中相关度较高的网页通过人们的自主选择获得了不同程度的加权,使得其在返回结果中的排名得到提升,更容易被检索到,提高了查准率。

This paper adopted the main idea of ant colony algorithm to improve the PageRank algorithm. Categorize the Web page according to the position of it in original sequence, compute the clicked number of the Web page acquired through the keyword search, and use the existing transfer function to operate the acquired result. The result was used as a factor to evaluate the degree of correlation between keywords and Web pages, so as to update the weight of the Web page, then obtain a new sort result. The simulation experiments show that this method is through people＇s subjective choice, so closely associated with the returned results page, their weight has been added with various degrees of growth. Its ranking in the returned results would be enhanced and more easily retrieved and the precision ratio was improved.