Towards anonymous semantic trajectories

In recent years, spatio-temporal and moving objects databases have gained consi-derable interest, due to the diffusion of mobile devices and of new applications, where the discovery of consumable, concise, and applicable knowledge is the key step. Recent advances in spatio-temporal data analysis focused on the semantic aspects of the movement data, thus leading to the definition of semantic trajectory concept. However, the analysis of this kind of data can compromise the privacy of users because the location data allows inferences which may help an attacker to discovery personal and sensitive information, like habits and preferences of individuals. In this paper we briefly present an approach for the generalization of semantic tra-jectories that can be adopted for obtaining datasets satisfying the k-anonymity property; specifically, this method exploits ontologies to realize a framework for publishing semantic trajectories while preserving privacy of the tracked users. We show that this generalization method is able to preserve the semantic tagging obtained by the analysis of the resulting dataset.