贝叶斯垃圾邮件过滤系统的设计与实现

DESIGN AND IMPLEMENTATION OF BAYESIAN SPAM FILTERING SYSTEM

分析了贝叶斯垃圾邮件过滤器的工作原理、分词、特征提取等相关技术,研究了Java Mail以及邮件相关标准和协议,设计了基于贝叶斯的垃圾邮件过滤系统;实现了服务器端的训练集管理器和客户端的邮件分类器、简易的邮件收发系统三大功能模块;在对邮件的处理中增加了人工复检和特征串匹配降噪的二次处理来完善过滤系统。

This paper analyzes the working principle of Bayesian spam filter, word segmentation, feature extraction and other related technologies, JavaMail and mail related standards and protocols are studied, A Bayesian spam filtering system is designed; The sys- tem implements the server training set manager and the client＇s mail classifier, simple mail transceiver system three functional mod- ules ; Two times in the processing of the mail is added to the manual testing and character string matching noise to improve the filte- ring system.