CBR IN THE CONTEXT OF ALBM: THE USE OF CASE-BASED REASONING TO ENHANCE CONVENTIONAL KNOWLEDGE-BASED SYSTEMS

CASE-BASED REASONING (CBR) REPRESENTS A POWERFUL NEW PARADIGM FOR BUILDING EXPERT SYSTEMS THAT PROMISES TO OFFER MANY ADVANTAGES IN REDUCING THE TIME AND COST OF KNOWLEDGE-ENGINEERING, IN SIMPLIFYING SYSTEM MODIFICATION, AND IN AUTOMATING DOMAINS THAT RESIST REPRESENTATION IN RULES OR FRAMES. COGNITIVE SYSTEMS PROPOSES TO EXPLORE THE PRACTICALITY OF THESE AND OTHER PROMISED ADVANTAGES FOR CBR IN THE CONTEXT OF AIR LAND BATTLE MANAGEMENT (ALBM), A EXPERT SYSTEM THAT WILL HELP TO MAKE THE PERFORMANCE ISSUES FOR ALTERNATIVE TECHNOLOGIES CONCRETE. COGNITIVE HAS ALREADY DEVELOPED A TOOL FOR BUILDING CBR SYSTEMS AND HAS USED THAT TOOL TO DEVELOP A CBR BATTLE PLANNING APPLICATION WITH SOME PARALLELS TO ALBM. COGNITIVE WILL DRAW UPON THESE APPLICATIONS AND TOOLS TO IDENTIFY POTENTIAL AREAS FOR CBR ENHANCEMENT TO ALBM. SUCH ENHANCEMENTS MAY EXIST AS SUBSYSTEMS FOR HANDLING RESISTANT PROBLEMS, DUPLICATE MODULES FOR ENHANCEMENTS IN PROCESSING TIME OR COMPONENTS FOR EXCEPTION PROCESSING. COGNITIVE WILL USE THIS EVALUATION TO PROVIDE A BLUEPRINT FOR PHASE II CBR ENHANCEMENTS TO ALBM. IN ADDITION, COGNITIVE WILL ASSESS THE AVAILABILITY OF DATA FOR CASE BASES. IF DATA IS AVAILABLE IN A DATABASE FORMAT WHICH MAY BE READILY DOWN-LOADED INTO A CASE BASE, COGNITIVE WILL DEVELOP A PROTOTYPE TO BE USED TO DEVELOP COMPARATIVE CBR/RULE-BASED PERFORMANCE DATA.