改进的Apriori算法在IDS中的应用  

The Application of Improved Apriori in IDS

作　　者：安德智[1] 

机构地区：[1]甘肃政法学院计算机科学学院,兰州730070

出　　处：《河北理工大学学报（自然科学版）》2011年第1期95-99,71,共6页Journal of Hebei Polytechnic University:Social Science Edition

基　　金：甘肃政法学院重点科研资助项目2008XZD028

摘　　要：在入侵检测研究领域中,提高检测模型的检测率并降低误报率是一个重要的研究课题。提出了一种针对网络入侵检测事务流日志数据库的关联规则挖掘改进算法,它采用事务压缩和属性压缩相结合,解决了当前主流关联规则算法应用到入侵检测过程中存在的多遍扫描、大量无效规则和算法复杂度过高等问题。实验结果表明,文中所提出的方法在规则生成和对网络异常情况的检测方面都显示出比较好的性能,提高了系统效率,使其更适用于入侵检测系统。In the field of intrusion detection research, it is of great importance to improve the testing rate of the detection model and decrease the false alert. The article suggests an improved algorithm of association rule mining, which combines transaction reduction and attribute reduction to solve repeated scan, dead rules and greatly complicated algorithm that occur when the mainstream association rule algorithm is applied in the field. The experiments indicate that the relevant way works well in testing rule production and network anomaly while improving the system efficiency.

关 键 词：入侵检测 关联规则 APRIORI算法 事务压缩 属性压缩 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]