A guarded cross-site mining architecture of systems security information

Abstract

This research effort provides a framework for a system that can securely fuse the intelligence from these sources, while completing the computing and communication in an efficient manner. We develop an architecture for a guarded cross-site mining system; this is designed to extract patterns and attack/intrusion indications as possible and utilize parallel processing of all relevant information available, while protecting sensitive information. This solution harnesses the power of the distributed computing environment by applying expert systems locally before aggregating data (instead of processing all at once at one central location). Furthermore, previously established theories for privacy preserving data mining may now be utilized for information assurance purposes.