Análise das relações socioespaciais dos crimes de furto e roubo na subprefeitura de Itaquera, São Paulo/SP - Brasil: uma abordagem utilizando inteligência artificial

Abstract

Introduction: Theft and robbery crimes, categorized as property-related offenses, pose threats to public safety and the quality of life in cities. Studies aim to comprehend their causes and patterns, with the intention of enhancing urban security policies. Theories point to individual, sociodemographic, and spatial variables as explanatory factors. Due to the unique characteristics of each locality, the topic demands further research, especially within spatial criminology, which investigates the relationship between crimes and socio-spatial elements. Identifying such elements and their connection to criminal patterns is crucial for devising policies for urban violence prevention and intervention. Objective: This study analyzed the spatial distribution patterns of theft and robbery crimes and their associations with urban socio-spatial attributes in the subprefecture of Itaquera, Eastern Zone of São Paulo/SP – Brazil. The selection of Itaquera was driven by the high incidence of property-related crimes in the subprefecture and the presence of elements associated with these crimes in the literature. Methodology: This is a quantitative exploratory study, based on documentary analysis and secondary data of theft and robbery crimes that occurred between 2017 and 2021, along with sociodemographic indicators. The analyses were supported by spatial analysis techniques, Geographic Information System (GIS), Artificial Intelligence (AI), and Data Mining (DM). Results: With the assistance of AI and DM, the approach identified contextual patterns with good accuracies (above 70%) that explain relationships between sociodemographic variables and occurrences of theft and robbery. In comparison to other studies, the findings revealed both convergences and divergences. Additionally, patterns associated with robberies were unveiled based on victim profiles. Conclusion: By deepening the understanding of how spatial patterns of theft and robbery crimes are influenced by sociodemographic characteristics, this approach provides a valuable scientific contribution, which can be applied in other socio-spatial contexts and guide future urban security policies. Furthermore, this study contributes to achieving the Sustainable Development Goals (SDGs) established by the UN, including Reduced Inequality (SDG 10), Sustainable Cities and Communities (SDG 11), and Peace, Justice, and Strong Institutions (SDG 16).