支持向量机在反垃圾邮件过滤中的应用  

Support vector machine application on filter against spam

作　　者：邹汉斌[1] 雷红艳[1] 邓卫红[2] 

机构地区：[1]湖南文理学院计算机科学与技术系,湖南常德415000 [2]张家界航空工业职业技术学院信息工程系,湖南张家界427000

出　　处：《计算机工程与设计》2007年第9期2015-2017,共3页Computer Engineering and Design

基　　金：湖南省教育厅基金项目(06C585)

摘　　要：垃圾邮件对计算机系统的安全和人们的生活造成了严重的威胁,反垃圾邮件问题已经成为的具有重要现实意义的研究课题。针对垃圾邮件过滤本质是分类问题,提出了一种基于服务器前端的反垃圾邮件过滤方法,它采用了改进的v支持向量机算法对邮件内容进行分类,过滤垃圾邮件。研究结果表明该方法与直接的支持向量机增量算法相比,提高了过滤的准确率,具有一定的应用价值。Spare has posed the serious threat to the computer system security and people＇s life, so the question against spam already became the important practical research topic. In view of which the essence of spam filter is the classify problem, one kind of filter method in front of the server against spam is proposed. It used the improvement v-support vector machine algorithm to carry on the classification to the mail content, to filter spam. The findings indicated that, this method and the direct support vector machine increase algorithm compared, enhanced the rate of accuracy which filtered, had the certain application value.

关 键 词：垃圾邮件 邮件过滤 分类 支持向量机 增量算法 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]