As personal data have been the new oil of the digital era, there is a growing trend perceiving personal data as a commodity. Although some people are willing to trade their personal data for money, they might still expect limited individual privacy loss, and the maximum tolerable privacy loss varies with each individual. In this paper, we propose a framework that enables individuals to trade their location data streams under personalized privacy loss, which can be bounded in w successive time points. However, the introduction of such personalized bounds of individual privacy loss over streaming data raises several technical challenges in the aspects of budget allocation, utility estimation of personalized differentially private mechanism, and arbitrage-free pricing. To deal with those challenges, we modularize three key modules in our framework and propose arbitrage-free trading mechanisms by combining instances of the modules. Finally, our experiments verify the effectiveness of the proposed mechanisms.
View on arXiv