Study On Graph Partition In Trajectory Publishing Privacy-Preserving

In the era of Internet of things,Location based services(LBS),develop rapid-ly becuase the ubiquity of cheap embedded sensors and smart phones.With the increase of LBS applications and the passage of time,more and more application servers store a large amount of trajectory information.The original trajectory data released directly without the privacy preserving may reveal the user's privacy,so the privacy protection of the trajectory becomes extremely important.In this paper,we propose three privacy-preserving approaches for trajectory publication based on k-anonymous and graph partition,specially:Firstly,we propose a privacy-preserving approach for trajectory publication with considering the uncertainty in trajectory because the inherent.inaccuracy of data acquisition equipments may lead to uncertain.The correlation between two trajectories are computed according to the temporal overlap similarity,the trajectory direction similarity and the distance between trajectories with uncertainty.The trajectory set.is mapped into a topological graph,and the k anonymous set is constructed based on k-node partition.Users can adjust the weight of the three parameters flexibly in order to achieve the tradeoff between the level of privacy protection and data utility.Secondly,we propose a trajectories k-anonymity privacy protection method over road network.Original trajectories data are transformed into simplified trajectories,represented as the Voronoi region identifications sequence that the original trajec-tories passed by.The Voronoi region is result of reasonable map partition based on Voronoi diagram.Trajectories similarities are computed by the edit distance among the Voronoi region identifications sequence.A greedy cluster algorithm based on k-node partition is proposed to construct k anonymous sets for privacy preserving.Thirdly,we propose an approach to protect stay points privacy by generalize or suppress stay points over road network based on Voronoi diagram.We partition the map to build secure anonymous area,then identify the significant stay points In order to protect stay points,we suppress or publish those stay points according to which region they belong to.