Banca de DEFESA: ARTHUR PIMENTEL GOMES DE SOUZA

Uma banca de DEFESA de MESTRADO foi cadastrada pelo programa.
STUDENT : ARTHUR PIMENTEL GOMES DE SOUZA
DATE: 18/03/2022
TIME: 14:00
LOCAL: Online
TITLE:
KEY WORDS:

Multicriteria decision aiding; Statistical analysis; GIS-MCDA; COVID-19; Public
security

PAGES: 163
BIG AREA: Engenharias
AREA: Engenharia de Produção
SUMMARY:

The state of Pernambuco has been distinguished in recent years due to the drop in crime rates
after the implementation of the Pacto pela Vida program. However, the advent of the COVID-
19 pandemic in 2020 generated a serious socio-economic crisis worldwide. Restrictive
measures reduced the movement of people on the streets and, consequently, affected criminal
activity. In this sense, the present study seeks to build a decision support model to investigate
the vulnerability of a region in the city of Recife, Brazil, to robberies. For this, statistical and
spatial techniques were combined on COVID-19 spatial propagation data and georeferenced
data on robberies in order to provide decision-makers with information during the construction
of their preferences. At first, the epidemiological situation in the city of Recife was explored
based on the evolution of the disease from month to month in view of government restrictions.
So the study was restricted to a phase of the beginning of the pandemic (April to July 2020) for
which there is availability of georeferenced data at the neighborhood level. Initially, the
evolution of the clusters of cases and of case-fatality rate on ten specific dates was examined,
accompanied by the exploration of the socioeconomic characteristics of these locations through
quartile analysis. In addition, a regression analysis presented how essential services and
socioeconomic characteristics were able to explain the number of cases reported over the ten
selected dates. In a second moment, the study of robberies was delimited to the neighborhoods
of Boa Viagem and Pina (Recife, Brazil). The evolution of robbery clusters in these
neighborhoods in different temporal divisions before and during the pandemic was then
investigated, as well as the socioeconomic characteristics of these areas. This was followed by
an exploration of 37 socioeconomic factors and commercial facilities based on quintile analysis,
Pearson correlation, and spatial correlation. Afterwards, five specific months representative of
different phases of the pandemic were selected to conduct a regression analysis, identifying the
main predictors. Finally, a multi-criteria model was built to obtain the classification of the
census sectors in Boa Viagem and Pina in terms of vulnerability to robberies.

BANKING MEMBERS:
Externo à Instituição - SILVIO HAMACHER
Presidente - 1510962 - CAROLINE MARIA DE MIRANDA MOTA
Interno - 1549303 - CRISTIANO ALEXANDRE VIRGINIO CAVALCANTE