MID: An innovative model for intrusion detection by mining maximal frequent patterns
Hui Wang, Chuanxiang Ma, Hongjun Zhang · 2012
Intrusion detection is a very important topic in dependable computing. Intrusion detection system has become a vital part in network security systems with wide spread use of computer networks. It has been the recent research focus and trend to apply various kinds of data mining techniques in IDS for discovering new types of attacks efficiently, but it is still in its infancy. The most difficult part is their poor performance and accuracy. This paper presents an innovative model, called MID, that counts maximal frequent patterns for detecting intrusions, needless to count all association rules, can significantly improve the accuracy and performance of an IDS. The experimental results show that MID is efficient and accurate for the attacks that occur intensively in a short period of time.