基于图和信息熵的入侵检测分类器构造

A Construction Method of Intrusion Detection Classifier Based on Graph and Information Entropy

针对传统的Snort入侵检测引擎所使用二维表逐一比较效率低下的问题,该文采用基于图的数据挖掘技术挖掘规则、运用改进后的信息熵方法选择分类属性,构造入侵检测分类器。通过实验证明,该方法是有效可行的。

In order to solve the question of traditional Snort intrusion detection engine used the two-dimensional table matching inefficient one by one, the paper applies graph-based data mining technology , uses improved information entropy methods to choose the classification attributes, thus eonstruct a intrusion detection classifier. The experiment results show that the method is feasible and effective.