Implementation of K-Means Clustering in Recognizing Crime Hotspots and Traffic Issues Through GIS

Abstract

The challenge of accurately identifying instances of crime and traffic issues has rendered the precise localization thereof difficult, thereby impeding the populace's access to information concerning areas of high risk and safety. Employing a Geographic Information System (GIS)-based mapping system utilizing the K-means clustering method, spatial data pertaining to crime and traffic concerns are grouped. The primary objective is to aid in the identification of high-risk areas concerning crime and traffic matters. The methodology employed in this study revolves around the application of the K-means clustering method to categorize spatial data relevant to crime and traffic issues. K-means clustering represents a non-hierarchical cluster analysis technique designed to partition data into multiple groups based on spatial similarities. Research findings elucidate that through the utilization of the K-means clustering method, three distinct sets of clusters predicated upon the intensity of crime and traffic issues emerge. Consequently, from these clustering outcomes, districts and specific locales falling within each cluster, denoted as moderately vulnerable (C1), vulnerable (C2), and highly vulnerable (C3), can be delineated. This system is poised to furnish recommendations to pertinent authorities for addressing areas exhibiting heightened intensity levels while concurrently facilitating the generation of reports and dissemination of information to the public via a dedicated website pertaining to areas at elevated risk of crime and traffic issues.