Privacy Preserving in Sequential Data Publishing

Abstract

Nowadays, sequentially captured data, i.e., trajectory datasets, are growing attention for data analytics because of the massive use of location-based service devices. Sharing such trajectory datasets with the researcher is vital for location data mining, such as traffic jams and people’s places of interest. Before the publication of trajectory datasets, enough privacy must be provided such that the confidentiality of each trajectory is preserved. In our solution, we apply space shifting operation on spatiotemporal points to minimize the record linkage attack by an adversary who knows some locations that are visited by a trajectory. We proposed a trajectory spatiotemporal set algorithm that gathers k number of various trajectories locations. Finally, we applied generalization operation on each set to achieve k-anonymity property where each spatiotemporal trajectory location is identical from k-1 other trajectories within a set.An extensive experiment result shows that enough trajectory privacy is preserved while publishing sequential trajectory data.