On Preserving Sensitive Information of Multiple Aspect Trajectories In-House

Abstract

Nowadays, human trajectories are enriched with semantic information having multiple aspects, such as by using background geographic information, by user-provided data via location-based social media, as well as by data coming from various kind of sensing devices. This new type of multiple aspects representation of personal movements as sequences of places visited by a person during his/her movement poses even greater privacy violation threats. This paper provides the blueprint of a semantic-aware Moving Object Database (MOD) engine for privacy-aware sharing of such enriched mobility data and introduces an attack prevention mechanism where all potential privacy breaches that may occur when answering a query, are prevented through an auditing methodology. Towards enhancing the user-friendliness of our approach, we propose a mechanism whose objective is to modify the user queries that cannot be answered due to possible privacy violation, to 'similar' queries that can be answered without exposing sensitive information.