期刊论文

A Mail Classification and Recognition Model Based on Neural Network

中文

南华大学学报(自然科学版)

1673-0062

2008

22

2

80-83

10.19431/j.cnki.1673-0062.2008.02.019

基金项目：湖南省教育厅资助项目（05C486）.

本校为第一机构

本文综合分析了垃圾邮件和合法邮件的特征,对邮件结构字段信息和邮件正文信息加以离散和特征化处理,提取出7个特征属性来表示成向量代表电子邮件,采用BP神经网络来构造邮件分类识别器.经测试证明,本文提出的基于BP神经网络的邮件分类模型是可行和有效的,并具有良好的效果和智能性、自学习性.

The paper synthetically analyzed the characteristics of spam and legal emails, dispersed and characterized emails then distilled seven characters to express emails on vectors. Neural network is used for the model. The result of the experiment shows that mail classification and recognition model based on neural network works well and has better result an...