A Ranking Method for Location-based Categorical Data in Smart Cities

• Alice A. F. Menezes UFAM
• Carlos M. S. Figueiredo UEA

Resumo

With the arising of Smart Cities and the amount of useful information that is provided by citizens, government, and companies nowadays, several interdisciplinary studies in Urban Computing became achievable. The literature present applications and analysis of cities dynamics, citizens mobility, and others. In this work, we propose a ranking method to detect different functional regions through virtual sensors using location-based categorical data of multiple sources. Also, we computed a diversity index according to the categories contained in regions of a study area. As a result, we found patterns associated with the characteristics of these regions, that influenced the inferring of functional regions by the ranking method in temporal analyzes. We evaluated our method using real-world datasets from Foursquare (200,339 check-ins) and NYPD Motor Vehicle Collision (209,908 occurrences) in New York City in 2017. The aim was to evaluate the outcomes of the proposed method concerning different types of categorical data. Also, we presented a case study considering the integration of these data. The generated results can be used in applications related to urban planning and can benefit citizens, entrepreneurs, and the government.