A Louvain-Based Approach to Discover Communities and Spatial Relations in a Homicide Knowledge Graph

Homicide is a major worldwide issue, and Mexico City is no exception, as illustrated by official data. In this study, we employ the Louvain algorithm that prioritizes modularity value to identify unique clusters within the extensive Mexico City homicide dataset stored in a knowledge graph. Our analysis unveils clusters housing varied murder occurrences that share connections within those same clusters. Spatial analysis within specific geostatistical divisions known as AGEBs (Área Geoestadística Básica) provides insights into spatial relationships and segregation regarding homicides, contributing to a deeper comprehension of Mexico City’s homicide landscape. Moreover, the Louvain algorithm’s capacity to identify optimal communities within the graph enhances its relevance for urban analysis by uncovering data relationships. These insights enhance urban planning and security strategies in Mexico City. This research advances the understanding of homicides and highlights the potential of advanced graph analysis techniques. Leveraging Louvain’s capabilities, we strive to inform data-driven interventions and enhance safety and security measures, ultimately promoting a safer environment in Mexico City.