基于结合内容特征的TrustRank算法改进  

Improvement of TrustRank algorithm based on combination of content features

作　　者：刘阳[1,2] 张化祥[1,2] 

机构地区：[1]山东师范大学信息科学与工程学院,山东济南250014 [2]山东省分布式计算机软件新技术重点实验室,山东济南250014

出　　处：《计算机工程与设计》2013年第4期1276-1279,1330,共5页Computer Engineering and Design

基　　金：国家自然科学基金项目(61170145);教育部高等学校博士点专项基金项目(20113704110001);山东省自然科学基金和科技攻关计划基金项目(ZR2010FM021;2008B0026;2010G0020115)

摘　　要：垃圾网页在利益的驱使下采用作弊手段欺骗搜索引擎获得更高的排名,干扰了用户对信息的获取。通过分析网页内容特征及其分布,提出了结合内容特征信息与TrustRank算法的方法对垃圾网页进行检测。实验结果表明,结合了内容特征信息的TrustRank算法能够有效的检测出垃圾网页。Driven by the benefit, web spam deceives search engines get high ranking, which disturbs users to obtain information normally. Detecting web spare is one of the major challenges faced by search engines. According to analysis of content features of web pages and their distributions, a new method is proposed to detect the spare pages, and the method takes into account the content feature information in TrustRank algorithm. Experimental results show that TrustRank can effectively detect spam pages with the help of the web page content feature information.

关 键 词：垃圾网页 搜索引擎 网页排名 信任指数 内容特征 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]