一种基于C4.5决策树的Web页面分类算法  

A C4.5 Decision Tree Based Algorithm for Web Pages Categorization

作　　者：曹薇[1] 张乃洲[2] 

机构地区：[1]武汉职业技术学院计算机学院,湖北武汉430074 [2]湖北大学知行学院,湖北武汉430011

出　　处：《计算机系统应用》2010年第10期195-198,共4页Computer Systems & Applications

摘　　要：WEB文本自动分类在很多方面都有着重要的应用,如信息检索,新闻分类等。决策树算法是一种简单并且广泛使用的分类方法,具有很多优点如:分类精度高,分类速度快等。主要研究了运用C4.5决策树构建Web页面分类器的基本方法和过程,并提出了一个基于C4.5决策树的Web页面分类器的框架。在此基础上实现了一个运用于网络爬虫的Web页面分类器,实验结果表明该算法是非常有效的。Web text categorization can be applied to many domains such as information retrieval, news categorization, etc. Decision tree algorithm is a simple method for categorization and has been used extensively. This paper investigates the basic method and process to build a web classifier by means of C4.5 decision tree, which has various merits such as high categorization precision, high categorization speed, etc. Moreover, this paper proposes a C4.5 decision tree based frame of web pages classifier, and implements it on a web crawler. The experimental results show that this algorithm is highly effective.

关 键 词：WEB文本分类 C4.5决策树 信息论 信息增益率 网络爬虫 

分 类 号：TP393.092[自动化与计算机技术—计算机应用技术]