Network intrusion detection system based on data mining

Abstract

Given increasing trend of computer networks containing highly crucial information, protecting against attacks and intrusion is of great importance. The present paper aims to design a network intrusion detection system (NIDS). Therefore, various methods adopted one of which is data mining. Data mining process performs in different ways. One way is ant colony optimization algorithm (Ant-Miner). Network intrusion detection system accommodates normal behavior patterns and detects intrusion based on the amount of deviation from normal behavior. The system relies on finding anomalies in network users’ behaviors involving two phases of learning and intrusion detection. The proposed system is based on data KDD99 extracted from University of California database and the test results revealed good performance in comparison to C5, CVM, and Cup Winner methods