Um comparativo do algoritmo de DBSCAN em ambientes de movimento livre e em rede de ruas

Abstract

In recent years, as a result of technological advancement, a large volume of data is generated every day by different technologies. Among these technologies are navigation data of vehicles, provided by GPS sensors. The volume of data grows considerably and it is currently a field of study to different kinds of entities. Thus, there is a need for effective solutions for the extraction of information. The objective of this work is the implementation of DBSCAN algorithm, which uses clustering technique, to compare two different environments to analyze geospatial data generated by vehicles in Beijing, China. The first environment, free movement, uses Euclidean distance and the second environment uses a Road Network, which considers the actual physical properties of the map. Tests were conducted to compare the two environments, with the effectiveness and shortcomings of each one, the execution speed, and accuracy of results.