Paper Contents
Abstract
People's data are frequently seen to be dispersed around many companies, yet when combined, these data can produce valuable insights. Even with certain security measures, data fusion from several data hosting sites may compromise user privacy. This work examines a data-analytic platform that integrates and analyzes user location and health data, which are kept in two different clouds, to identify infectious disease geographic hotspots using the Kulldorff scan statistic. We investigate the privacy risks associated with this platform, which employs a key-oblivious inner product encryption (KOIPE) approach to guarantee that the honest-but-curious (HbC) entity is only exposed to coarse-grained statistical data. Utilizing a game-theoretic strategy, we safeguard user privacy against the intended inference attack by providing incentives for users to create clusters that are anonymous and ensure quantitative privacy. We show the effectiveness of our system in terms of design overhead and privacy level through comprehensive simulations based on real-life datasets. Index Terms: game theory, secure multi-party computing, Bayesian inference, public health, and Kulldorff scan statistic.
Copyright
Copyright © 2024 U. Vara Prasad, Ch. Sai Krupa. This is an open access article distributed under the Creative Commons Attribution License.
