基于免疫克隆选择算法的垃圾网页检测  

WEB SPAM DETECTION BASED ON IMMUNE CLONAL SELECTION ALGORITHM

作　　者：杨凡[1] 朱焱[1] 唐寿洪 

机构地区：[1]西南交通大学信息科学与技术学院,四川成都610031

出　　处：《计算机应用与软件》2015年第6期20-23,29,共5页Computer Applications and Software

基　　金：四川省学术带头人培养基金项目(x8000912371309)

摘　　要：垃圾网页是指一些网页通过不正当的手段来误导搜索引擎,使网页获得高于其应有的排名,从而获得更多的访问量。它不仅降低了网页的质量,同时也导致了严重的Web信息安全问题。传统的垃圾网页检测通常使用经典的机器学习方法包括贝叶斯算法、SVM、C4.5等,这些算法对垃圾网页的检测有一定的效果。在前人的研究基础上提出一种基于免疫克隆选择的垃圾网页检测方法。利用人工免疫系统的自学习及自适应能力来检测利用新作弊技术的垃圾网页,并与广泛用于垃圾网页检测的贝叶斯算法对比。实验表明该方法能有效、可靠地检测出垃圾网页。Web spam refers to those Web pages which mislead search engines through improper means to get higher ranking than they deserve,so they may get more access.The Web spamming not only reduces the quality of Web page,but also leads to serious security problems of Web information.Traditional Web spam detection usually uses classical machine learning approaches including Na？ve Bayes, SVM,C4.5,etc.,they are effective to some extent for detecting Web spam.We propose a novel immune clonal selection-based Web spam detection according to previous studies.It uses adaptive and self-learning ability of artificial immune system to detect the Web spam utilising new fraud technology,and is compared with Na？ve Bayes which is widely used to detect Web spam.Experiment reveals that this approach is able to effectively and reliably detect Web spam.

关 键 词：搜索引擎 垃圾网页 克隆选择 分类 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]