Uncovering Crime Patterns: A Geospatial Clustering Approach

Abstract

According to the National Institute of Statistics and Geography (INEGI), the state experienced a notable increase in criminal incidents during the year preceding its September 2024 report on victimization and public perception of security, surpassing figures from the previous three years. Considering this trend, public security institutions must formulate strategic responses based on historical data that reveal when, where, and how crimes occur. In this regard, this article analyzes crime patterns in Tamaulipas, Mexico, as one of the most pressing social challenges. Utilizing the K-Means clustering algorithm, this study identifies latent patterns within crime data to classify neighborhoods in the state’s principal municipalities according to their levels of criminal activity. The analysis yields a detailed and well-defined segmentation of these clusters, incorporating both criminal and socioeconomic variables. The primary contribution of this research is the development of an information system featuring an interactive map that enables geospatial visualization of crime clusters at the neighborhood level. This tool is designed to assist public security authorities in the effective allocation of resources and the formulation of targeted crime prevention strategies.