This paper was originally published as a Technical Report No. 213, April 1, 2005 for the Ivan E. Seidenberg School of Computer Science and Information Systems (CSIS) Department.

Document Type



We discuss the design of an intrusion detection and resonse system (idrs) generator. This design involves a belief fuser, a belief tree classifier, and a memoryless fuzzy incident responder. The firm's security policy,its current risk profile,and training data constitute input streams. The system is designed to produce an incident response that security officers feasibly adopt to improve the firm's risk position as indicated in the corporate security policy. We do not present a prototype for the idrs generator but we provide sufficient details on the credal and pignistic schemes for the fuser and the classifier, needed to develop the idrs generator.