LAUSR

People use smart transportation systems to move around in smart cities, producing a massive amount of valuable mobility data. Although this characteristic enables the development of many intelligent applications, it can expose users to privacy threats. Location privacy is an issue addressed in many mobility contexts, in which there is a privacy concern. Currently, there are some proposals to tackle this problem, and some questions naturally arise: Are these proposals suitable for a dynamic environment, such as smart mobility? What are the impacts of mobility on privacy? In this article, we answer these questions to explore location privacy in smart mobility considering open and online data, one of the fundamental pillars of smart city platforms. We have evidenced the hypothesis that mobility can impact privacy approaches of anonymization (mix zones) and obfuscation (GEO-I) in the context of smart mobility. For this, we performed experiments to characterize and find similarities in the statistical distributions extracted from two stay points metrics, which operate as substrates to build location privacy protection mechanisms. We use an accuracy metric to quantify the datasets’ distributions that matched each other. We conducted a comprehensive evaluation of seven real datasets of mono or multimodal mobility. The results showed that the stay point count metric reached 100% and 83.3% accuracy for coarse-grained (person and vehicle) and fine-grained (bus, taxi, and person) data. Additionally, we show a similarity between distributions for the same vehicle type for mono and multimodal datasets. Results suggest that privacy has a high dependence on mobility in different granularity levels.