HEURISTIC, OR EXPERIENCE-BASED DECISIONMAKING IS MORE RELIABLE FOR FIGHTING CRIME THAN FORMAL TESTS OF STATISTICAL SIGNIFICANCE THROUGH SOPHISTICATED DATA ANALYSIS. HOWEVER, DATA ANALYSIS CAN BE VALUABLE IN VERIFYING OR DENYING EXPERIENCE-BASED DECISIONS. DETERMINISTIC CRIME PREVENTION, WHICH ASSUMES THAT ENVIRONMENT CAUSES AND CONDITIONS CRIMINAL BEHAVIOR, COMBATS CRIME BY CHANGING THE ENVIRONMENT. DATA SUPPORT COLLECTS INFORMATION ABOUT CRIMINALS AND THEIR ENVIRONMENTS AND TESTS THEORIES WITH FORMAL RESEARCH DESIGNS. THIS APPROACH CAN SUCCEED ONLY IF DECISIONMKERS HAVE CONTROL OVER THE ENVIRONMENT, OBTAIN LEGAL AND SOCIAL SUPPORT, AND PROVIDE SUFFICIENT DATA TO LINK CRIME TO ITS CAUSES. THE ARTIFICAL INTELLIGENCE APPROACH ASSUMES THAT CRIME PREVENTION DECISIONS ARE BASED ON INCOMPLETE DATA IN PARTICULAR ENVIRONMENTS--ACCEPTABLE SOLUTIONS ARE FOUND THROUGH FORMAL BUT CHANGEABLE DECISION RULES INTERACTING WITH EXPERIENCED HUMAN JUDGMENTS. THE URBAN ANALYSIS PROJECT WAS DEVELOPED TO ASSIST LOCAL GOVERNMENTS IN EXPLORING DATA ANALYTIC AND HEURISTIC APPLICATIONS TO COMMUNITY CRIME PREVENTION. CRIME IS UNDERSTOOD HERE AS THE RESULT OF COMMUNITY DEFICIENCIES; COMMUNITIES ARE AIDED IN DEFINING THEIR PROBLEMS AND MARSHALLING THEIR RESOURCES. (PAP)