基于序列模式的网络入侵检测系统

Network intrusion detection system based on sequential patterns

针对入侵检测的特点将数据挖掘技术应用于网络入侵检测系统,阐述了网络入侵检测系统的设计原理及其实现。系统的数据挖掘模块应用了序列模式挖掘中的GSP算法,并对其进行了改进,引入了主属性及兴趣度。实验表明,优化后的算法可以有效地提高检测的准确率,使系统的性能获得提升。

The paper applies data mining technology to network intrusion detection system（NIDS） according to the characteristics of NIDS. The design principles and the implementation of the NIDS are illustrated. The main attribute and interest measure are introduced to improve the （GSP） algorithm, which is then applied in the data mining module of the system. The results of experiments show that the precision and performance of the NIDS are improved by the optimized algorithm.