Advanced search
Start date
Betweenand

Visual analysis and engineering of urban features for crime prediction in São Paulo City

Grant number: 19/10560-0
Support type:Scholarships in Brazil - Post-Doctorate
Effective date (Start): August 01, 2019
Effective date (End): July 31, 2020
Field of knowledge:Applied Social Sciences - Demography
Principal Investigator:Sergio França Adorno de Abreu
Grantee:Erick Mauricio Gómez Nieto
Home Institution: Faculdade de Filosofia, Letras e Ciências Humanas (FFLCH). Universidade de São Paulo (USP). São Paulo , SP, Brazil
Associated research grant:13/07923-7 - Center of the Study of Violence - NEV/USP, AP.CEPID

Abstract

São Paulo is the largest city in South America, as well as one of most diverse in types of crime. During last decades, its criminality rates have down considerably, probably by a combination of factors as a decline in the number of young people, a decrease of the unemployment rate, changes in law enforcement practices and strict control on firearms. However, the reasons to explain such a phenomenon are not completely clear. Looking for an understanding of these causal and consequence relationships is crucial to support government making-decision and ensure urban safety.This work proposes the development of a methodology for analyzing data from multiple contexts (e.g. socio-economic, crime records, environmental infrastructure) produced by São Paulo city with the purpose of identifying relationships/patterns/trends and including them into a predictive model that allows us to anticipate crime behavior. The partnership of the Centro de Ciências Matemáticas Aplicadas à Industria (CEMEAI-USP) and Núcleo de Estudos da Violência (NEV-USP), both at Universidade de Sao Paulo is the fundamental key for this study due to the interaction with experts in mathematical and computational sciences in one side, and in sociology and urban geography in the other, bring us the opportunity to assess our results in both fields of study.The effectiveness and usefulness of the proposed methodology will be demonstrated in case studies involving real data and validated by domain experts and by the capability to identify phenomena described in the literature.