基于用户反馈的个性化垃圾邮件过滤方法  

Personalized spam filtering method based on users' feedback

作　　者：黄国伟[1] 刘云霞[1] 陈志[2] 

机构地区：[1]深圳信息职业技术学院,广东深圳518172 [2]天津工业大学,天津300387

出　　处：《电子设计工程》2014年第15期53-56,共4页Electronic Design Engineering

基　　金：广东省自然科学基金资助项目(S2011040006119)(S2012010008964);深圳市科技计划项目(JCYJ20120615103057639)

摘　　要：如何在传统垃圾邮件过滤技术基础上实现垃圾邮件个性化过滤是垃圾邮件过滤领域的重要课题。提出一种基于用户反馈的个性化垃圾邮件过滤方法,一方面将用户反馈应用于邮件分类特征的更新,提取用户个性化邮件分类标准;另一方面,将全局邮件分类标准和用户个性化分类标准综合应用于朴素贝叶斯分类过程,实现用户邮件个性化分类。仿真实验结果表明,在用户邮件分类标准存在差异的环境下,基于用户反馈的个性化垃圾邮件过滤方法能够有效提升传统垃圾邮件过滤技术的邮件分类效果。How to realize the personalized spam filtering based on traditional filtering methods is an important problem in the area of spam filtering. A personalized spam filtering method based on users’ feedback was proposed. On one hand, the feedbacks are used to extract the global and the user’s personalized standard on the e-mail classification, respectively;on the other hand, these two standards are integrated to realize the personalized spam filtering in the proposed method. The simulation results show that the proposed method is able to improve the performance of the traditional spam filtering method in e-mail classification.

关 键 词：垃圾邮件 邮件分类标准 用户反馈 个性化邮件分类 贝叶斯模型 

分 类 号：TN99[电子电信—信号与信息处理]