Classificação de atividades do usuário do transporte público por meio de smartphones

Abstract

Mobility should be a universal right. However, the concepts of city and mobility are not completely in harmony. In Brazil, much of the workers’ time is wasted in chaotic city traffic. Public transport is far from being efficient and the public authorities fall short of this challenge. Therefore, several works propose technological solutions in search of urban mobility with a focus on real-time travel time prediction. To achieve this goal, it is necessary to deepen the study of classification of activities related to the means of locomotion used. Currently, there are solutions with great potential using mobile networks and smartphones through several algorithms that enable the classification with accuracy. However, even so, Brazilian cities are far from intelligent public transport. The objective of this work is to deepen the knowledge about the classification of activities related to public transport, through the J48 algorithm implemented in Weka, using accelerometer data, in order to preserve user privacy. Keywords: Classification. J48.Public transport.