距离函数分类法在垃圾邮件过滤中的应用  

Application of categorization based on distance function in spam filtering

作　　者：林琛[1] 李弼程[1] 

机构地区：[1]解放军信息工程大学信息工程学院,河南郑州450002

出　　处：《计算机工程与设计》2007年第2期322-323,447,共3页Computer Engineering and Design

基　　金：河南省教育厅基金项目(sp200303099)

摘　　要：为了得到实用性强的垃圾邮件过滤方法,将距离函数分类法首次引入到垃圾邮件过滤中。在通用邮件语料库上进行测试,并与目前过滤性能较好的KNN算法进行比较,实验结果显示距离函数分类法中的类中心向量法不适合用于垃圾邮件的过滤,而类重心向量法在保持较高过滤性能的同时,具有训练和过滤速度快的优点,是一种理想实用的垃圾邮件过滤方法。In order to get applicable performance method of spam filtering, categorization based on distance function is firstly applied to filter spare. It is tested on e-mail corpus and compared with KNN method that is good method in spam filtering. Experimental result show categorization based on category center vector is bad method for spam filtering and categorization based on category centroid method not only is better than KNN in filtering performance, but also the speed of training and filtering is high. It is a good and useful method for spam filtering.

关 键 词：垃圾邮件 分类 距离函数 类重心向量 垃圾邮件过滤 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]