基于内容与链接特征的中文垃圾网页分类  

Chinese Web spam Identification through Content and Hyperlinks

作　　者：刘卫红[1] 方卫东[1] 董守斌[1] 张凌[1] 

机构地区：[1]华南理工大学计算机学院,广州510641

出　　处：《微计算机信息》2010年第9期6-8,共3页Control & Automation

摘　　要：随着搜索引擎使用的日益普及,web作弊已成为搜索引擎面临的一个重大挑战。国内外研究人员从基于内容,基于链接等方面提出了许多反web作弊的技术,这些技术一定程度上能有效地检测垃圾网页。本文在前人研究基础上提出了一种结合网页内容和链接方面的特征,采用机器学习对中文垃圾网页进行分类检测的方法。实验结果表明,该方法能有效地对中文垃圾网页分类。Web spam has become one of the severest problems for today＇s network systems .Overseas researchers have proposed many techniques to combat web spoofing based on the content of and/or the links in a Web page.These techniques have been proved to be useful in detecting spam Web pages.However, most of these techniques are only applicable to pages in foreign languages, and in particular English.So far, little research is done to deal with Web spam in Chinese.Based on related works, we propose a novel Chinese Web spam detection method using machine learning technologies.Experimental results show that the proposed method can detect Chinese Web spam effectively.

关 键 词：网页作弊 搜索引擎 机器学习 

分 类 号：TP393.098[自动化与计算机技术—计算机应用技术]