Research On Sensitivity Recognizing,privacy Preserving For Trajectory Data Publication

The trajectory data of users can reflect their interests and preferences.If there is no anonymity,these private data cannot be directly published.Based on the user's trajectory data,the attacker can infer other sensitive location information of the user based on the spatiotemporal correlation of part of the user's location,leading to the disclosure of the user's privacy.At present,most privacy protection methods in trajectory data publication either regard all location information as sensitive information or distinguish sensitive locations solely from location labels or access frequency,to improve the utility of data.However,different locations have different sensitivities for users,if only consider location tags or access frequency,there will be over-protection and under-protection,on the one hand,it is impossible to protect users' privacy,and on the other hand,it reduces the utility of the published trajectory data.Therefore,this paper studies how to accurately identify the sensitive location of users and the corresponding privacy protection scheme in the publication of trajectory data.In this paper,we propose a precise identificatio n and adaptive anonymity method for sensitive information based on knowledge graph,PSR&PPM?KG(Precise Sensitivity Recognizing,Privacy Preserving,Knowledge Graph-based Method for Trajectory Data Publication),which can accurately identify sensitive locations of different users on the trajectory and provide a level privacy security guarantee.The main research work of this paper is as follows:(1)This paper summarizes and analyzes the methods of identifying the sensitive locations on the user's trajectory and the privacy protection methods of a trajectory data release in offline application scenarios.It points out that the existing methods are not accurate enough to identify the user's sensitive locations on the trajectory data,so that there are problems such as low data utility and serious information loss in solving the privacy protection of trajectory data publication,and the existing technologies and methods are elaborated and analyzed in detail.(2)To solve the problem that sensitive location identification in user trajectory is not accurate enough,this paper constructs a trajectory knowledge graph that integrates user attributes,location labels and their relation.Then an accurate sensitive location recognition algorithm is designed based on the knowledge graph.K nowledge is obtained by extracting association rules from the knowledge graph,which can accurately identify the personalized sensitive locations of the user granularity level.(3)Based on the sensitive location identification method proposed above,an adaptive anonymization method is proposed,which has privacy protection for accurately identifying sensitive locations.This paper analyzes performance from two aspects of privacy and information loss,proves that the method proposed in this paper provides sufficient privacy protection for the publication of trajectory data and improves the utility of the published trajectory data.(4)This paper conducts experimental tests for the above methods on the commonly used datasets O ldenburg and Movielens and evaluates the performance from three aspects:,and.Experimental results show that the PSR&PPM?KG algorithm has more advantages in data utility.