Implementação de algoritmos para análise de similaridade de trajetória na biblioteca PyMove

Abstract

With the growth in the volume of trajectory data, numerous studies have been done to detect new insights and understand about movement patterns, classifications, predictions and the behavior of objects in geographic space. There is a large number of research on trajectory data. However, it is noticed that there are not many tools capable of helping researchers to extract from this data all the information they can offer. Among the various mechanisms used in this service, the measure of similarity between trajectories demonstrates great efficiency in the analysis of trajectory data. For this reason, this work implements operations to recover trajectories based on similarity in time and space, making available to any person the operations developed, both for use and for their improvement. For the analysis of operations, several types of data sets are used. It also presents some difficulties of these recoveries and contexts that can change the way we can base them, in addition to making comparisons with others already developed.