基于Bayes的一种改良垃圾邮件过滤模型  

A New Model for Improved Spam E-mail Filter Based on Bayes

作　　者：龚伟[1] 

机构地区：[1]涪陵师范学院计算机科学系

出　　处：《微计算机信息》2007年第3期104-106,共3页Control & Automation

基　　金：国家教育部"春晖"计划(Z2005-1-55003)

摘　　要：文章首先分析了垃圾邮件的产生机理,介绍了目前比较常见的几种垃圾邮件过滤技术,然后从朴素贝叶斯的理论依据出发,针对当前应用于重要商业领域的垃圾邮件过滤系统的不足,设计了一种应用多级邮件策略的新模型,并通过实验比较证明新模型的应用在一定程度上提高了垃圾邮件过滤系统的查全率和查准率。The article has first analyzed the spam production mechanism, introduced quite common several kind of spam filters tech-nology at present, then embarks from native Bayes theory,in view of current spam filters system the insufficiency applies to the im-portant commercial domain, has designed one kind of application classification mail strategy new model, and compared the certificate new model through the experiment the application to enhance the spam in the certain degree to filter the system the recall and the accuracy ratio.

关 键 词：垃圾邮件 过滤 实时黑名单 朴素贝叶斯 邮件分级 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]