PageRank的改进算法——调整阻尼因子  

PageRank Improved Algorithm-Adjustment Damping Factor

作　　者：邵晶晶[1] 李波[1] 刘汉平[1] 

机构地区：[1]华中师范大学数学与统计学学院,湖北武汉430079

出　　处：《应用数学》2008年第S1期57-61,共5页Mathematica Applicata

基　　金：湖北省自然科学基金项目(2007ABA337)

摘　　要：Google将PageRank定义成某个非周期不可约Markov转移概率矩阵的平稳分布,于是对PageRank算法的改进所得到的矩阵一定要是非周期不可约Markov转移概率矩阵,结合RageRank算法和林共进修正算法思想,以及修正算法存在的问题,本文给出了改进算法,并通过简单试验对改进算法进行调整,调整后的改进算既满足Google的初衷又解决其算法的问题,也没有增加算法的复杂度.PageRank was defined as the stationary distribution of some aperiodic irreducible Markov transition probability matrix,therefore the matrix which obtains to the PageRank algorithm's improved must certainly be the aperiodic irreducible Markov transfer probability matrix,unifies the PageRank algorithm and the revision algorithm thoughts of Lin Gongjin and the problems of the revision algorithm,this article has given the improved algorithm,and makes the adjustment through the simple experiment to the improved algorithm,after the adjustment improved algorithm not only satisfy the original intention of Google algorithm and resolve their problems,without increasing the complexity of the algorithm.

关 键 词：PAGERANK算法 Markov转移概率矩阵 平稳分布 改进算法 

分 类 号：O211.6[理学—概率论与数理统计]