入侵检测系统中关联规则的挖掘

The Data Mining Of Association Rules in the Intrusion Detection System

在分析几种现有关联规则算法的基础上,针对网络数据的具体情况,引入关键属性、参考属性以及属性相对支持度的约束,对关联规则挖掘算法FP_Growth进行了扩展,解决了基本关联规则挖掘算法中产生大量无用模式的问题,从而帮助系统发掘出更有意义的模式,提高了系统挖掘的执行效率和规则库的准确度.

This paper solves the problem that there is abundant useless modes produced by basic association rules mining, which involves the key attributes, referenced attributes and the restriction of attributes＇ relative support. It extends the algorithm of association rules according to the specific condition of the network data. This extension has assisted the system mine more meaningful patterns and raised the execution efficiency and accuracy of rules sets.