数据挖掘技术在邮件病毒检测中的应用研究  

Application Study of Data Mining Technique in E-Mail Virus Detection

作　　者：徐宗龙[1] 潘宏侠[1] 谭谈[2] 

机构地区：[1]中北大学自动控制系,山西太原030051 [2]中北大学通信与信息系统,山西太原030051

出　　处：《机械管理开发》2005年第6期70-72,共3页Mechanical Management and Development

摘　　要：针对电子邮件病毒危害大和更新速度快的特点,以及传统的使用手工编制的启发知识在新病毒上的软弱,研究了数据挖掘的归纳学习和分类方法,利用数据挖掘方法产生新邮件病毒的知识来解决传统方法的缺陷。通过贝叶斯方法建立了检测知识的识别模型和自学习模型,构建了检测系统的结构框架,并作出了深入的讨论。Aiming at the characteristics that E-mail virus is badly harmful and its renewal speed is highly quick, and traditionally using the heuristic knowledge that is compiled manually has the weakness in detecting new virus, the paper studies the induction learning and sorting method of data mining and takes advantage of data mining methods to generate the detection knowledge of new E-Mail virus, which can resolve the defect of traditional method. In this paper, the recognition model and self-learning model of detection knowledge have been built by bayes method, and the structural frame of the detecting system has been constructed, and makes some detailed discussions.

关 键 词：病毒检测 电子邮件 数据挖掘 贝叶斯 

分 类 号：TP309.5[自动化与计算机技术—计算机系统结构]