基于云计算的Pagerank算法的改进  

An improved Pagerank algorithm based on cloud computing

作　　者：郑晶[1] 

机构地区：[1]福建江夏学院电子信息科学学院,福建福州350108

出　　处：《福州大学学报（自然科学版）》2014年第1期45-49,共5页Journal of Fuzhou University(Natural Science Edition)

基　　金：国家自然科学基金资助项目(30671680);国家科技型中小企业技术创新基金资助项目(11C26213502126);福建省教育厅科技资助项目(JA11269);福建江夏学院青年资助项目(2011C005)

摘　　要：针对Pagerank算法在Web结构挖掘中存在的需要大量迭代的问题,提出一种新的方法.该方法通过对原始Pagerank值的计算公式进行改进,降低了迭代次数.实验表明,在云计算环境下,新方法减少了网络通信和访问HDFS的消耗,在时间花费上优于传统的Pagerank算法.With the advent of the era of cloud computing, it is a new important research topic to discuss the problem of the web mining based the cloud computing. A new method is proposed to solve the large number of iterations problems in the Web structure mining for the Pagerank algorithm. Through improving the formula of the original pagerank value, it reduces the number of iterations. The experiments show that this method reduces the network traffic and the consumption of accessing HDFS in the cloud computing enviroment, and it is superior to the original Pagerank algorithm in the time consumption.

关 键 词：云计算 WEB结构挖掘 PAGERANK MAPREDUCE 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]