Developing a Law Enforcement Framework Using a Data Mining Approach: A Case Study on Criminal Justice in Gindeberet Woreda Court, West Shoa Zone

Abstract

Crime is a person‟s behavioral disorder that result instability of citizens‟ normal life in all circumstances; social, economic, environmental, and political means. Today, security is the world‟s major concerns and becomes a primary agenda along the world. Because, the issue grows in intensity and complexity, following its dramatic peak of existence. Analysis of such large collection of criminal data, to make appropriate decision becomes difficult and complex. To overcome such problems available technologies are being recommended to use. Here the researcher tried to model the integration of DM, to recommend and assist the stakeholders in decision making related to crime incidents. To generate rules, decision tree and rule induction algorithms were selected as these techniques are widely used to represent real world problems in the form of rule. PART rule induction algorithm recorded best and it used to acquire knowledge. Because, PART rule and PROLOG follows contradictory algorithm, forward chaining and backward chaining respectively, the researcher develop a model that change PART rule to PROLOG supported syntax form for implementation. The performance of the system was evaluated using the test cases prepared for this purpose. Twenty-five test cases were prepared for system performance test and provided to domain experts. Users were trained for conducting user acceptance test and were asked to fill test forms provided. Generally, the system has scored 70% overall performances which is a promising result.