基于转移概率的PageRank算法研究  

Transition Probability-based PageRank Algorithm

作　　者：李吉平[1] 吴陈[1] 曾庆军[1] 

机构地区：[1]江苏科技大学电子信息学院,镇江212003

出　　处：《科学技术与工程》2008年第8期2071-2074,共4页Science Technology and Engineering

基　　金：江苏省科技计划项目(BE2006086)资助

摘　　要：PageRank算法是目前一种基于网络链接的比较成功的网页排序算法。首先对PageRank算法做了简单介绍,然后从数学的角度分析,PageRank可以被看作是一个马尔可夫随机游走模型,基于此模型对经典PageRank算法中的转移概率做了改进,根据网页的原PageRank值在链入网页的所有链出网页PageRank值总和所占比例定义转移概率。实验表明,改进后的算法在计算出网页的PageRank值,更具合理性。PageRank is one of the successful algorithms which based on the links of the network at present. At the beginning, a simple introduction for the classic PageRank algorithm is given, then, from the point of mathematics, PageRank can be seen as a Markov random walk model, the ambulant probability changes from the same in the classic PageRank algorithm to base on the value of old PageRank of the website in the improved PageRank algo- rithm. The experiments show that the Rank values which are calculated with the improved algorithm are more rea- sonable than before.

关 键 词：PAGERANK 链接分析 马尔可夫随机游走 转移概率 

分 类 号：TP391.3[自动化与计算机技术—计算机应用技术]