Trajectory Data Publication Through Individualized Sensitive Stay Location Anonymization

Abstract

The abundance of GPS embedded devices accumulates trajectories in an excessive scale and it is enriched with personal information. While publishing traces for research activities, we must ensure to publish the anonymized trajectory in order to prevent the disclosure of individual privacy. During anonymization, we need to anonymize the stay locations where the user considered it as most sensitive instead of anonymizing all locations. For finding and extracting the most sensitive stay locations, we adopt a new method by considering the individual spatial and temporal factors using SSLF function. This combines the stay points within a threshold distance to form stay locations and also anonymize these locations in a stay zone using the generalization SSLA approach. The proposed model is tested with a real-world dataset and it guarantees a better trade-off between privacy and utility compared with other models of same nature.