电子邮件智能分类系统的设计

Design of Email Intelligent Classifier

互联网时代 ,人类的几种主要通讯方式中 ,电子邮件是既快捷又经济的方式 .对当前人们往来的大量电子邮件 ,一般在电脑网络中 ,使用基于关键字比较之分类系统的过滤器对邮件进行分类 .这种传统方式存在缺乏灵活性 ,不能有效地处理概括性描述复杂性问题等缺点 .使用效果不甚理想 .本文在传统规则分类方法的基础上引入机器学习的知识 ,设计了一个新的电子邮件智能分类系统 ,较好地解决了概括性的分类问题 。

In the Internet Age, Email is one of the fastest and the most economical methods of communication. The large number of emails, by which people communicate with each other nowadays, are usually classified by the filters, which use the regular classifier based on keyword-compared. However, this traditional method has some shortcomings, such as lack of flexibility, incapability of describing the complicated problems summarily and effectively. Therefore, its effectiveness is not very satisfactory. This paper introduces the knowledge of machine learning on the base of the traditional regular classification. A new intelligent email classifier is designed to solve the summary classified problems well and demonstrate the feasibility of the system through experiments.