Abstract: As network traffic becomes more and more heavy and intrusion method begins to be more various, current ID techniques are often unable to precisely or effectively describe intrusion, normalcy and anomaly. Aimed at this kind of shortcoming,this paper describes a data mining approach for building intrusion detection models. It also discusses some related techniques and the results of the experiment.

摘要： 当前入侵检测技术在网络流量剧增和入侵手段不断变化的情况下，往往无法精确、高效地描述入侵行为、正常行为和异常行为。针对这一不足，提出一种应用数据挖掘技术进行入侵检测模型的构建的方法，并介绍了模型中应用的相关技术以及一些实验结果。

关键词: 入侵检测, 数据挖掘, 分类, 序列分析

CLC Number: 

• TP393.08