Application of Fuzzy Association Rules in Intrusion Detection

Abstract

Intrusion detection system (IDS) is based on data mining technology in this paper. Association rule mining which is a method of data mining will make the boundaries of intervals hard. It will increase the information loss. In this paper, a novel framework based on data mining techniques is proposed for designing IDS. In this framework, the classification engine, which is actually the core of the IDS, uses fuzzy association rules for building classifiers. Particularly, the fuzzy association rule sets are exploited as descriptive models of different classes. Generally, the proposed approach outperforms other methods, especially in terms of false positive rate.