Location data record privacy protection based on differential privacy mechanism

Abstract

Now many location data applications have facilitated people’s daily life. However, publishing location data may divulge individual sensitive information. Currently many existing privacy protection schemes cannot provide the balance of utility and protection. Furthermore, as the records about location data may be discrete in database, some existing privacy protection schemes are difficult to protect location data information in data mining. In this paper, our works mainly focus on providing a framework for the privacy protection of location data mining. We propose a location data record privacy protection scheme based on differential privacy mechanism, which employs the structure of multi-level query tree to query and publish location data on database. Our proposed location data privacy protection scheme may discover the relationship of location data from database and protect location data mined. As accessing location preference of user may be related to private (sensitive) location, it is very important to protect highly frequent accessing location data when location data are mined. So, our proposed scheme provides a mechanism to protect highly frequent accessing location data (or accessing location preference of user) by distorting accessing frequencies. In the proposed scheme, we first construct the structure of multi-level query tree from database, then we make double processes of selecting data according to accessing frequencies by the exponential mechanism and one process of adding noises to accessing frequencies by the Laplace’s mechanism on the multi-level query tree. Compared with the other schemes, the experiments show the data availability of the proposed scheme is higher and the privacy protection of the scheme is effective.