A differentially private method for reward-based spatial crowdsourcing

Abstract

The popularity of mobile devices such as smart phones and tablets has led to a growing use of spatial crowdsourcing in recent years. However, current solution requires the workers send their locations to a centralized server, which leads to a privacy threat. One of the key challenges of spatial crowdsourcing is to maximize the number of assigned tasks with workers’ location privacy preserved. In this paper, we focus on the reward-based spatial crowdsourcing and propose a two-stage method which consists of constructing a differentially private contour plot followed by task assignment with optimized-reward allocation. Experiments on real dataset demonstrate the availability of the proposed method.