| 引用本文: | 崔超,杨威,张宪忠,张志军.贝叶斯网络在过滤垃圾邮件算法中的应用研究[J].哈尔滨工业大学学报,2011,43(11):145.DOI:10.11918/j.issn.0367-6234.2011.11.030 |
| CUI Chao,YANG Wei,ZHANG Xian-zhong,ZHANG Zhi-jun.Bayesian application study on arithmetic for filtering junk e-mail[J].Journal of Harbin Institute of Technology,2011,43(11):145.DOI:10.11918/j.issn.0367-6234.2011.11.030 |
|
| 摘要: |
| 为在用户数据流中删除垃圾邮件,研究了具有自我学习能力的自适应邮件过滤系统.在对正常和垃圾2类邮件误分类成本分析的基础上,利用概率性的学习方法创建满足过滤任务需要的过滤器,且讨论使用邮件域名特征变量进行特定邮件过滤并设计了过滤器,最后对实际邮件组进行操作,验证了算法的可靠性. |
| 关键词: 贝叶斯理论方法 概率 特征变量 邮件过滤 |
| DOI：10.11918/j.issn.0367-6234.2011.11.030 |
| 分类号:TP393.098 |
| 基金项目: |
|
| Bayesian application study on arithmetic for filtering junk e-mail |
|
CUI Chao1, YANG Wei2, ZHANG Xian-zhong1, ZHANG Zhi-jun3
|
|
1.College of Applied Technology,Qiqihar University,161006 Qiqihar,China;2.Qiqihar Information Center,161006 Qiqihar,China;3.Qiqihar information Technology Institute,161006 Qiqihar,China)
|
| Abstract: |
| To delete spams from User data stream,an adaptive self-learning spam filtering system has been studied and presented.On the basis of cost analysis mistakenly classified of normal and spam e-mail,a filter to meet the requirements of filtering tasks with learning methods of probabilistic is created,and the use of mail domain name characteristic variables for a particular e-mail filtering is studied.Finally,experiments verify the reliability of the algorithm. |
| Key words: bayesian theory probability feature-variable filtering e-mail |
