Técnicas de visualização de dados para auxílio ao processo de tomada de decisão na polícia militar no município de Caicó/RN

Abstract

Data visualization is understood as a data representation science in graphical format. It allows the decision makers to analize statistics in a visual way, making possible them identify patterns in the data and getting more accurate understanding about difficults found on them. The present work is a exploratory research where were established methods, parameters and techniques for data visualization. The data used in research came from a militar police criminal database in the city of Caicó-RN. The main objective of this research it is comprehend the criminal growth in city with the usage of data visualization integrated with the KDD process over the criminal ocurrences, the base hold 37.554 entries of ocurrences recorded in the period of 2008 and 2014. Were applied visualization techiques like graphs, hotspots mapping, data comparison of criminal ocurrences and time series analysis using the following technologies from Google: Google Maps API 3, Google Maps Geocoding API and Google Charts. With usability of visualization techniques in hotspots maps it was possible to see the yearly criminal growth in the city, just like the accompaniment of criminal expansion in the year of 2013, through out the twelve months. With the usability of graphs was possible to obtain statistics to identify which crimes most happened with greater rate and what periods of seven year most there were occurrences records, also was possible to do a parallel criminal comparison over the municipal festive periods predetermined to the analysis. It's concluded that this research was satisfatory, bringing relevants results to the police decision making in the city, although not were addressed all possible visual tecniques and the database had many invalids records to hotspot mapping.