AN AI BASED DECISION SUPPORT SYSTEM FOR SELECTION OF RAMP PROGRAM PARTS CANDIDATES

THE OBJECTIVES OF THIS PROPOSAL ARE TWOFOLD. FIRST IS TO DEVELOP A DETAILED CONCEPTUAL DESIGN OF AN OPERATIONAL DECISION SUPPORT SYSTEM (DSS) FOR THE SELECTION OF RAMP PROGRAM PARTS CANDIDATES, AND SECONDLY, TO DEVELOP A PROTOTYPE OF THE MORE CRITICAL ELEMENTS OF THE DSS SYSTEM. THE ENVISIONED SYSTEM WILL CONSIST OF AUTOMATED INPUT AND OUTPUT COMMUNICATION LINKS, A DECISION SUPPORT SYSTEM RESIDENT ON APPROPRIATE HARDWARE AND INCLUDING A MENU-DRIVEN INTERFACE, DECISION LOGICS, AND OUTPUT STORAGE CAPABILITIES. EXPERT SYSTEMS PROGRAMMING WILL BE USED IN DEVELOPING SELECTION CRITERIA AND IN EVALUATING CANDIDATE PARTS. IN ADDITION, EXPERT SYSTEMS PROGRAMMING AND NATURAL LANGUAGE PROCESSING CAPABILITIES WILL BE CONSIDERED FOR USE IN EVALUATION OF PARTS CANDIDATE DATA AND FOR REACHING CONCLUSIONS ON THE BASIS OF INCOMPLETE DATA. SEVERAL ASPECTS OF AN OPERATIONAL SYSTEM WILL BE STUDIED FOR DETERMINATION OF THE MOST APPROPRIATE APPROACH. SPECIFIC TOPICS TO BE ADDRESSED INCLUDE 1) RULE SELECTION METHODOLOGY, INCLUDING USE OF "INDUCTIVE" AI TOOLS, 2) SELECTION CRITERIA AND POSSIBLE USE OF USER-SET RULES 3) "THRESHOLD" VS. "SPECIFIED" CANDIDATE QUANTITIES FOR SELECTION, 4) THE NEED FOR A THREAT SPECIFIC ANALYSIS CAPABILITY FOR SPECIFIC SYSTEMS AS OPPOSED TO ANALYSIS OF THE FULL UNIVERSE OF PARTS, 5) OPTIONS RELATING TO USE OF EXISTING DATA ONLY VS. MANUAL INTERVENTION WITH MISSING DATA, AND 6) THE NEED FOR A PERIODIC REVIEW AND COMPARISON CAPABILITY.