Incorporating Background Knowledge in Privacy Preserving Geolocation: An Access Control Approach

Privacy is an important issue that should be considered before releasing individual's data for research purpose. Recent studies proved that even in the form of low resolution maps, researchers were able to determine the relative error between the true house locations and the location determined by ego-referencing the published map. This problem is more complicated by the fact that, in addition to the released point maps or data in general, one may also has access to other resources (e.g., published maps on surveys, public records, voter lists, and social networks ), which known as Background knowledge. Spatio-temporal data refers to time-stamped location information, which is; in our setup can be collected using Geolocation technique. We employ location as a distance measure from which we can estimate the adversary’s background knowledge. Obviously, people from the same location tend to have more background knowledge about quasi-identifiers (QI) than people from other locations. Based on this rationale, we designed an extended Role-based access control “LPT-RBAC”, so that we limit one’s access to information based on Role-based access policy and preserve individual's privacy.