Abstract:With the development of mobile devices and location based service (LBS) in mobile internet,a large number of spatial data based on location were produced.By publishing and analyzing,spatial data can not only provide convenience for the personal life and enterprises production such as transportation navigation and mobile advertising,but also provide services for the government decision such as disaster emergency response.But,if they were not protected when publishing,there are serious privacy leakages because spatial data imply abundant personal information.Therefore,it is the key problem in spatial data opening and sharing that how to balance privacy protection and data availability.Aim to this problem,the idea of privacy protection based on data analysis rises.In this paper,to balance the privacy protection and the data availability in analyzing the opening and sharing spatial data set,a method of constructing users' interest regions including the space,time and people numbers is proposed.It is that the user interest regions are got by preliminary analyzing to spatial data series,and they will be used in privacy protection.The method has two steps.The first step is to get the individual interest regions,and the second is the public interest regions.In first step,we firstly preprocessed M mobile users' trajectory data with N sampling times,formalized them to the location matrix with M×N elements which implies the sequential relationship.Next,we clustered the row vector of each user in the location matrix according to the access frequencies.Finally,we merged and optimized the cluster result to get the individual interest regions in chronological order.The second step is based on the first.Firstly,we extracted the individual interest regions' location points of each user according to practical application.Next,we clustered the individual interest regions' location points of all users according to the distance.Finally,we scanned the index matrix with M× N elements which implies many information of the public interest regions related to different time scale to get the public interest regions.To verify the method is feasible and effective,we gave some theoretical analysis and reality testing.The number of user and the sampling rate is the main influence factors to the time and space complexity in our method.In the experiment of clustering parameters and individual interest regions' location points,we give many simulation experiment result to compare.By comparing the processing results from the public dataset and the entity data from Baidu map,we can see the interest region created with our method is same to the function region in life.In the application example,we used the public interest regions in trajectory privacy protection,and kept the statistical characteristics of data set.So,the application example gives the proof that the interest region with our method is the basic of trajectory privacy protection.In conclusion,we proposed a method of constructing users' interest regions,and they include the information of space,time and people numbers to meet the application requirement.Using this method,we can balance the privacy protection and the data availability in spatial data opening and sharing.

A User Study on the Capability of Three Geo-Based Features in Analyzing and Locating Trajectories

More Value, Less Privacy, How to Evaluate the Privacy Based on Trajectories Visualization & Analyzation

Exploring Multi-Scale Spatiotemporal Twitter User Mobility Patterns with A Visual-Analytics Approach

Experiments and Analyses of Anonymization Mechanisms for Trajectory Data Publishing.

Privacy-protected social media user trajectories calibration

A Survey and Experimental Study on Privacy-Preserving Trajectory Data Publishing

Research on Trajectory Privacy Preserving Method Based on Tra jectory Characteristics and Dynamic Proximity

Personalized Trajectory Privacy Protection Method Based on User-Requirement.

Personalized semantic trajectory privacy preservation through trajectory reconstruction

Understanding User Behavior Geospatially.

Inferring Demographics from Human Trajectories and Geographical Context.

Anonymization and De-Anonymization of Mobility Trajectories: Dissecting the Gaps Between Theory and Practice

Historical Trajectories Based Location Privacy Protection Query.

Utility-aware and Privacy-preserving Trajectory Synthesis Model that Resists Social Relationship Privacy Attacks

Trajectory Privacy-preserving through Significant Stays Protection

Spatiotemporal Mobility Based Trajectory Privacy-Preserving Algorithm in Location-Based Services

Here Is Not There: Measuring Entailment-Based Trajectory Similarity for Location-Privacy Protection and Beyond

Constructing Users' Interest Regions with Two Steps for Trajectory Privacy Protection

UrbanMotion: Visual Analysis of Metropolitan-Scale Sparse Trajectories

A Feature Based Method for Trajectory Dataset Segmentation and Profiling

User Identification across Asynchronous Mobility Trajectories.