Temporal and similarity analysis for predicting serial crimes

• Grant number: 24/08575-7
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: July 01, 2024
• End date: June 30, 2025
• Field of knowledge: Interdisciplinary Subjects
• Principal Investigator: João Luiz Becker
• Grantee: Paula Pécora de Barros
• Host Institution: Escola de Administração de Empresas de São Paulo (EAESP). Fundação Getúlio Vargas (FGV). São Paulo , SP, Brazil
• Associated research grant: 20/07019-2 - FGV Analytics - Public Security Policies Analytics Research Center, AP.NPOP

Abstract

We propose the creation of the FGV Center for Science Applied to Security with the support of the São Paulo StatePublic Security Secretariat and the IDB, responding to the call of NPOP-SP proposals made by FAPESP Science forDevelopment Program, on the theme of Smart Cities and Public Security. The objective is to establish a researchframework to promote the use of artificial intelligence systems to support decisions in the area of public security,providing the technical structure and adequate ambiance of articulation among the key actors in the sector andstimulating the development of evidence-based assessment tools to evaluate public policies. Three pillarsstructure the proposal, differentiating it from other initiatives: 1. Developing big data and AI tools, decisionsupport systems, and monitoring the results of decisions based upon them; 2. Evaluating public security policies;and 3. Structuring research networks and open innovation processes. The main initial challenges defined togetherwith the project partners are: (1) prediction of crimes to support operational planning, where research to findsolutions will involve the creation of computational/statistical models to predict crimes (predictive policing); (2)monitoring and preventing cargo theft, with research involving temporal and similarity analysis, forecastingprobabilities of events; (3) granular categorization of information contained in police reports filed by citizens (BOs)and incident reports from police officers (RAIAs), with research for solutions comprising text mining techniques;(4) prediction of serial crimes - the case of automatic teller machines, involving research in temporal andsimilarity analysis; (5) modeling the behavior of gangs and organized crime, with research focused on networkanalysis; (6) evaluation of security policies, conducting experiments, quasi-experiments and surveys to findsolutions; (7) evaluation of the different policing strategies adopted by the PMESP, with research encompassingthe development of a tool to support decisions; (8) effectiveness of policing programs to reduce criminality, whereresearch to find solutions will use ex-post evaluations of programs; (9) understanding whether the technologicaltools developed really have affected rates of criminality and security perception, including experimental evaluationof technology usage to combat crime; and (10) asymmetry between crime indices and security perception, withresearch involving development and collection of security perception metrics.