Professors Nélio Cacho and Allan Martins, researchers in the Data Analysis and Visualization work package, received a best paper award in the First IEEE International Conference on Smart City Innovations (IEEE SCI 2017), held on August 4-8, 2017 in Freemont, California, United States. The conference aims to promote a forum where multidisciplinary researchers and teams, industry professionals, and government agencies can exchange innovative ideas, discuss challenges, present research results and solutions, as well as project success cases.
The paper entitled Feature Engineering for Crime Hotspot Detection presents the work developed by the professors in collaboration with researchers from the Karlsruhe Institute of Technology (KIT), in Karlsruhe, Germany. As main contribution, the work presents a crime detection approach intended to promote improvements in public safety. Such an approach is mainly based on a classification algorithm using Machine Learning techniques along with both temporal and location information to make high-quality predictions on a given crime category. For evaluation purposes, data reporting crime occurrences in San Francisco (United States) and Natal (Brazil) were used. Obtained results revealed a correlation between features of these urban areas and crime-prone areas.