Spatial Cloaking Revisited: Distinguishing Information Leakage from Anonymity

  • Authors:

  • Kar Way Tan;Yimin Lin;Kyriakos Mouratidis

  • Affiliations:

  • School of Information Systems, Singapore Management University, Singapore 178902;School of Information Systems, Singapore Management University, Singapore 178902;School of Information Systems, Singapore Management University, Singapore 178902

  • Venue:
  • SSTD '09 Proceedings of the 11th International Symposium on Advances in Spatial and Temporal Databases
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

Location-based services (LBS) are receiving increasing popularity as they provide convenience to mobile users with on-demand information. The use of these services, however, poses privacy issues as the user locations and queries are exposed to untrusted LBSs. Spatial cloaking techniques provide privacy in the form of k -anonymity; i.e., they guarantee that the (location of the) querying user u is indistinguishable from at least k -1 others, where k is a parameter specified by u at query time. To achieve this, they form a group of k users, including u , and forward their minimum bounding rectangle (termed anonymizing spatial region , ASR) to the LBS. The rationale behind sending an ASR instead of the distinct k locations is that exact user positions (querying or not) should not be disclosed to the LBS. This results in large ASRs with considerable dead-space, and leads to unnecessary performance degradation. Additionally, there is no guarantee regarding the amount of location information that is actually revealed to the LBS. In this paper, we introduce the concept of information leakage in spatial cloaking. We provide measures of this leakage, and show how we can trade it for better performance in a tunable manner. The proposed methodology directly applies to centralized and decentralized cloaking models, and is readily deployable on existing systems.