摘要
分析了邮件循环的产生因素及传统的检测方法 ,建立了基于邮件会话中来源与目的的统计模型 ,研究了会话时间序列的统计特性 ,据此计算各来源目的对的循环概率值 。
This paper describes a novel approach to automatically detect looping messages. It first analyzes factors leading to mail loops and discusses several traditional solutions. Then constructs a statistical model based on each (source, destination) pair by in depth analysis of statistical properties of each pair's conversation time intervals. According to probability computing of each pair, we can easily detect which pair is in a message loop.
出处
《小型微型计算机系统》
CSCD
北大核心
2003年第6期969-971,共3页
Journal of Chinese Computer Systems
基金
科技型中小型企业技术创新基金 :( 0 1C2 62 14 2 10 3 89)资助