基于主题相关性和时间因素的改进PageRank算法  

Improved Page Rank Algorithm Based on Topic-relevance and Time

作　　者：耿瑞[1] 李石君[1] 尹为民[1] 

机构地区：[1]武汉大学计算机学院,湖北武汉430072

出　　处：《微电子学与计算机》2015年第8期158-162,共5页Microelectronics & Computer

基　　金：国家自然科学基金(61272109)

摘　　要：提出一种将主题相关性和网页新旧程度与PageRank算法结合,得到改进的PageRank算法TRTPR(Topic Relevance and Theme PageRank).本算法通过链接分析来评估网页的新旧和主题相关性,据此对网页排序.研究结果表明,改进的算法在基于时间和主题相关性上比传统的PageRank算法有更好的效果.An improved algorithm named TRTPR（Topic Relevance and Time PageRank）is given out,which combines topic relevance and pub-date with PageRank.The newness and oldness of web pages and their topic telation are judged according to the analysis of links.The research results show that the improved algorithm has better effect than PageRank algorithm on time factor and theme-relevance.

关 键 词：网页 发布时间 PAGERANK 主题相关性 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]