基于综合过滤技术的邮件过滤终端研究

摘要

To solve the problem that the spam filtering system on the server side hardly meets special demands of some certain groups for email filtration, after analysing the MIME protocol and their format features, we bring out a spam filtering system on the client side. The system is built up with three-layer filtering technology, i.e. , blacklist, feature filtering and Naive Bayesian algorithm; moreover the user feedback mechanism is adopted to reduce the false alarm rate. Experiment result shows that the novel system can meet users' personalised needs for filtration and improve the performance of spam interception.%为了解决基于服务端的垃圾邮件过滤系统难以满足某一群体邮件过滤的特殊要求,分析了Email邮件协议和邮件格式特征,设计与实现了一种客户端垃圾邮件过滤系统.系统采用黑白名单过滤、邮件特征过滤和贝叶斯分类相结合的三层过滤技术,并通过用户反馈机制降低误报率.实验结果显示,系统可满足用户对邮件过滤的个性需求,提高了垃圾邮件过滤的准确性.