Data Mining Technologies for Proactive Detection, Computer Forensics, and Active Response of Security Violations in Large Scale Information Systems
Small Business Information
500 West Cummings Park, Suite 3000, Woburn, MA, 01801
AbstractThe Phase II SBIR project will develop and evaluate a Methodology for Information Assurance in Large Scale Information Systems, centered on Monitoring, Detection and Response. Data Collection for off-line analysis and on-line monitoring will be coordinatedby COTS Network Management Systems (NMSs), enhanced by software modules for Applications and OS Management. The enabling technologies of Data Mining, Machine Learning, Statistical Pattern Recognition, Multivariate Time Series Analysis, and SystemIdentification will be used for selecting variables for monitoring, correlation of events, fusion of variables for optimal decision-making, fusion of decision-making modules, selection of detection thresholds, and extraction of Proactive Rules relatingTargets and Attackers. Two commercial products are envisioned: (i) a plug-in software module for standard COTS NMSs, capable of endowing NMS systems with alarming and active response and (ii) System Security Toolkit for companies who wish to engage insystem forensics after security violations have happened. Aprisma Management Technologies will provide consulting in network management and computer security, as well as support on implementation, marketing, commercialization and Phase III transition.North Carolina State University will provide a research testbed for design and evaluation. Prof. Sushil Jajodia (George Mason University) will provide consulting in Computer Security, Data Mining and Database Design.
* information listed above is at the time of submission.