A MULTIAGENT SYSTEM BASED AUTOMATED PROFILING MODEL TO AVOID INSIDER THREAT
AbstractThe proposed model builds and maintains the profile of all insiders to detect and avoid threat. Software agents are autonomously working to record all activities of the authenticated users. The profile is compared with the policy of the organization and insiderâ€™s profile is marked acceptable or suspicious. This will lead to a proper mechanism to protect organizations assets against threat. The model is generic, adaptable and works in most of the organizations. It follows the renowned agent standard of Foundation for Intelligent Physical Agents (FIPA) as agents built on other platforms can interact with the agents developed on the proposed model. The profiling models exist but agents based autonomous models have never been implemented and tested before. The testing and execution environment for Multiagent systems has been developed by ourselves and its compatibility with other agent models have also been checked.
A. Kamra, E. Bertino and G. Lebanon, Mechanisms for Database Intrusion Detection and Response, Proceedings of the 2nd SIGMD Ph.D Workshop on Innovative
Database Research (2008).