Constructing elastic distinguishability metrics for location privacy

Abstract

Abstract:With the increasing popularity of hand-held devices, location-based applications and services have access to accurate and real-time location information, raising serious privacy concerns for their users. The recently introduced notion of geo-indistinguishability tries to address this problem by adapting the well-known concept of differential privacy to the area of location-based systems. Although geo-indistinguishability presents various appealing aspects, it has the problem of treating space in a uniform way, imposing the addition of the same amount of noise everywhere on the map. In this paper we propose a novel elastic distinguishability metric that warps the geometrical distance, capturing the different degrees of density of each area. As a consequence, the obtained mechanism adapts the level of noise while achieving the same degree of privacy everywhere. We also show how such an elastic metric can easily incorporate the concept of a "geographic fence" that is commonly employed to protect the highly recurrent locations of a user, such as his home or work. We perform an extensive evaluation of our technique by building an elastic metric for Paris' wide metropolitan area, using semantic information from the OpenStreetMap database. We compare the resulting mechanism against the Planar Laplace mechanism satisfying standard geo-indistinguishability, using two real-world datasets from the Gowalla and Brightkite location-based social networks. The results show that the elastic mechanism adapts well to the semantics of each area, adjusting the noise as we move outside the city center, hence offering better overall privacy.

Efficient Utility Improvement for Location Privacy

“…Privacy under arbitrary metrics is studied in [39], while [18] proposes a method to construct a distinguishability metric taking into account the semantics of each location. In this paper we mostly assume d to be Euclidean, but also discuss mechanisms that allow the use of an arbitrary privacy metric.…”

“…The smaller exponent leads to a greater variance of the noise, hence the utility of this mechanism is the worse among those discussed in this section, with the advantage, on the other hand, of being very simple and at the same time applicable to any metric d. The exponential mechanism is used in [18] to achieve privacy wrt a constructed "elastic" metric, adapted to the semantics of each location.…”

“…[18] for its use in location privacy) and the tight-constraints mechanism [19] are applicable as well. The latter are applicable to any finite domain equipped with an arbitrary privacy metric.…”

Leveraging Hierarchical Representations for Preserving Privacy and Utility in Text

“…For example, a privacy model built using the Manhattan distance metric can be used to provide indistinguishability when the objective is to release the number of days from a reference point [14]. Similarly, the Euclidean distance on a 2d plane can be used to preserve privacy while releasing a user's longitude and latitude to mobile applications [15]. Finally, the Chebyshev distance can be adopted to to perturb the readings of smart meters thereby preserving privacy on what TV channels or movies are being watched [22].…”

“…To set the value of ε for a given task, we propose following the guidelines offered by [15] in the context of location privacy by providing appropriate reformulations. They suggest mapping ε to a desired radius of high protection within which, all points have the same distinguishability level.…”