基于特征构建的网络入侵检测研究

Research of Network Intrusion Detection Based on Feature Construction

入侵检测技术是保护计算机网络系统的一个重要技术。本文探讨使用数据挖掘技术构建入侵检测系统框架,基于网络连接数据挖掘出频繁模式构建的主要特征,并对有用的模式进行编码,构造分类器识别入侵行为,提出模式编码方法及特征构建算法,实现入侵检测框架的建立。

Intrusion detection is a very important technique to protect network systems. This article discuss to construct intrusion detection models using data mining techniques.Based on network connection data mining the main features of frequent pattern construction and encode useful patterns. Moreover, pattern encoding scheme and feature construction algorithm are proposed to create the frame of network intrusion detection system.