摘要
论文提出了一种基于邻近类别分类的邮件过滤系统模型,并介绍了系统涉及到的文本特征选择、贝叶斯分类算法等关键技术,最后给出了评价方法与实验结果。结果表明,该方法能够显著地提高系统对于垃圾邮件的查准率。
This paper proposes one model of Email Filtering System Based on Approximate Category Classification. In addition, some related key technologies, such as text feature selection and naive Bayes algorithm are introduced. Finally, this paper provides an assessment method and the results of experiments. It shows that this method can notably enhance the system precision toward mail spam.
出处
《信息安全与通信保密》
2006年第5期81-83,共3页
Information Security and Communications Privacy
基金
国家863高技术研究发展计划项目资助(编号:2003AA142160)
关键词
邻近类别分类
邮件过滤
特征选择
approximate category classification spam filtering feature selection
