Steven Harp, Ph.D.
Dr. Steven Alex Harp is a cognitive-scientist and statistician who has worked extensively on systems for decision support, automated reasoning, computer security, and agent-based systems. In addition to system design and experimentation, Dr. Harp has also been a very active software developer, translating his ideas into working prototypes for a wide range of commercial, industrial and government customers.
Information Assurance: Much of his recent work has focused on computer security in conjunction with the DARPA CyberPanel, OASIS, and SRS programs. This includes development of the Scyllarus system for correlating multiple diverse intrusion detection systems, and the development of a form of autonomy for critical systems, known as skeptical control. He created a novel mechanism for formulating SELinux policy that coordinates access restrictions over a complex network of clients and servers. He co-invented a system called BAMS for behavioral attack modeling using AI planning techniques. He devised a system called SCOAP for automatic synthesis of detection rules based on automatically generated attacker courses of action. In the area of Self Regenerative Systems, he worked on the SPDR system for recognizing malicious plans and identifying insider threats, the CORTEX system for learning to repel the most general forms of a novel attack, and the CSISM system, a coginitive architecture featuring rule-based automated defenses.
Background: During his time with Honeywell Labs, he was chief architect of the NIST ATP Abnormal Situation Management (ASM) program which resulted in the AEGIS system, a distributed, collaborative decision support system to assist industrial plant operators in abnormal situation management. The ASM results were initially applied to oil refineries, then to chemical plants and most recently to cyber security to identify network attacks. For his technical achievements on the ASM program, Dr. Harp was recognized with an H.W. Sweatt Award, Honeywell's highest engineering honor. Also for NIST, he helped to design and implement an agent-based hardware/software system to facilitate independent living for the elderly.