Use of Hybrid Data Mining in Identification of Crime Patterns and Trends in the Matatu Industry in Kenya

Abstract

The aim of this study was to propose an automatable technological framework that identifies crime and misconduct patterns and trends in the matatu industry using data mining techniques for intelligence led policing in Kenya. The objectives of the study include to propose a framework for intelligent transport management system with patterns and trends identification capabilities, enhance formulation of policy developments, implementations and government regulations for the transport sector in Kenya, design model system for testing the framework to ascertain its practicability and effectiveness and identify challenges of the transport sector in Kenya. This was an application research which made use of dummy data. The study established that it is possible to use artificial intelligence to manage the transport sector by use of a system that will not only help identify the patterns and trends of matatus' on Kenyan roads but to answer the why's associated with the trends to help come up with meaningful applicable practical solutions to enhance security and integrity in the transport sector in general. The study also unearthed challenges in relation to the implementation of the above. Combination of classification and association rules based data mining approach was utilized for this study due to its effectiveness in bringing out patterns and trends that are interlinked and related to each other.