Inner interruption discovery and defense system by using data mining

Abstract

Network security helps to prevent the network against the intruders from performing malicious activities. The security can be provided to the networks using firewalls, anti-virus software and scanners, cryptographic systems, Secure Socket Layer (SSL) and Intrusion Detection Systems (IDS). Authentication is the commonly used technique to protect the unauthorized users from the network. But, it is easy to compromise the login passwords using brute force attacks. The IDS and firewalls concentrate on the external attacks, while the internal attacks are not taken into account. In order to solve these issues, this paper proposes an Inner Interruption Discovery and Defense System (IIDDS) at the System Call (SC) level using data mining and forensic techniques. The user’s profiles are maintained and compared with the actual dataset using Hellinger distance. A hash function is applied on the incoming messages and they are summarized in the sketch dataset. The experimental results evaluate the proposed system in terms of accuracy and response time.