摘要
为了减少检测入侵防御系统的报警信息数量和误报率,设计了加权投票贝叶斯分类法,该分类法可以有效的将报警信息分类,减少重复报警。并使用具有高性能的报警关联算法与加权投票贝叶斯分类法有机结合,实现了针对入侵防御系统的报警分析器。通过对使用该报警分析器与未使用报警分析器的入侵防御系统进行比较实验,表明了该报警分析器可以有效的减少重复报警数量,降低误报率。
To decrease the number of alarm information and false positive ration of intrusion protection system, a new Bayesian classification algorithm which named weight polling Bayesian classification algorithm is designed. The new Bayesian algorithm is more accurate, and can reduce the number of alarm information remarkably. An alarm info analyzer is designed by integrating correlation algorithm with the Bayesian classification algorithm. The comparison experiments between the intrusion protection system with the alarm info analyzer and the intrusion protection system showed that the alarm info analyzer decrease the number of alarm info and false positive ration remarkably.
出处
《计算机工程与设计》
CSCD
北大核心
2008年第14期3620-3622,共3页
Computer Engineering and Design
基金
教育部"新世纪优秀人才支持计划"基金项目(NCET-04-0843)
关键词
入侵防御系统
关联
贝叶斯
加权
网络安全
intrusion protectionsystem
correlation
Bayesian
weight
network security
