Exploring Machine Learning Classifiers for Crime Data Analysis: Leveraging Computational and Environmental Criminology for Enhanced Insights

Abstract

In the recent era of constant technological development there is a remarkable shift in interconnected computational techniques in the sphere of criminology. Databases and tools like Digital Forensic, Fingerprinting, Facial Recognition, Video Analysis, Data Recovery are actively working and trying to set up societies free from crimes. This research seeks to propose a framework with the inclusion of the ML algorithms and criminology theories to analyse social indicators for crime and estimate crime trends to boost crime analysis capability. This is through understanding police-reported crime and the socio-demographic data hence affording crime analysis to be used in crime prevention. According to the findings, machine learning algorithms allowed the most promising performance, and more specifically, decision trees which performed the best at the feature selection stage and at analysis of the crime patterns. In addition, the work can be useful for improving the policy-making process for addressing and preventing crime by indicating such factors that stimulate and facilitate criminal activities.