计算机与现代化 ›› 2013, Vol. 1 ›› Issue (7): 43-046,.doi: 10.3969/j.issn.1006-2475.2013.07.011

• 算法设计与分析 • 上一篇    下一篇

基于社会网络的两阶段垃圾邮件过滤方法

黄国伟1,陈 志2   

  1. 1.深圳信息职业技术学院计算机学院,广东深圳518172;2.天津工业大学计算机科学与软件学院,天津300387
  • 收稿日期:2013-02-05 修回日期:1900-01-01 出版日期:2013-07-17 发布日期:2013-07-17

Double-stage Spam Filtering Method Based on Social Network

HUANG Guo-wei1, CHEN Zhi2   

  1. 1. Computer College, Shenzhen Institute of Information Technology, Shenzhen 518172, China;2. School of Computer Science & Software Engineering, Tianjin Polytechnic University, Tianjin 300387, China
  • Received:2013-02-05 Revised:1900-01-01 Online:2013-07-17 Published:2013-07-17

摘要: 如何对用户信誉进行动态更新以使之能准确反映用户的邮件发送行为,是基于用户信誉的垃圾邮件过滤方法需要解决的关键问题之一。以电子邮件用户之间存在的社会网络关系为基础,提出一种基于用户反馈的信誉评估机制,借助用户的反馈信息提升信誉机制对用户邮件发送行为特征变化的适应性。在此基础上,提出一种两阶段的垃圾邮件过滤方法,实现基于社会网络的用户信任与基于反馈的用户信誉在邮件分类中的有机结合。仿真实验结果表明,本文提出的过滤方法既能保证信誉机制对用户行为变化的良好适应性,又能获得较好的邮件分类效果。

关键词: 垃圾邮件, 邮件分类, 社会网络, 用户信任, 用户反馈, 用户信誉

Abstract: How to guarantee the accuracy of the reputation of E-mail users reflecting users’ behavior in sending mail is one of key problems in reputation-based spam filtering method. To handle this problem, a reputation scheme based on users’ feedback is proposed. The scheme recurs to the trust relations among users in social network and the information provided by users’ feedback in order to guarantee the adaptability to the dynamic changes in users’ behavior. Furthermore, a double-stage spam filtering method is proposed based on the reputation scheme, which integrates the trust among users and the reputation of users in the process of E-mail classification. The simulation results show that the proposed method is able to achieve both good adaptability to the changes of users’ behavior and good performance in E-mail classification.

Key words: spam, E-mail classification, social network, user’s trust, user’s feedback, user’s reputation