PMATE: A Privacy-Preserving Map Retrieval Task Assignment Scheme in Spatial Crowdsourcing

Abstract

Spatial crowdsourcing (SC) task assignment is to find the optimal worker for the task from abundant alternative workers based on the information of the task and workers, such as location, time, and ability. This information will undoubtedly reveal the privacy of both the task and workers. The disclosure of private information is a crucial issue constraining the development of SC. To this end, various privacy-preserving task assignments have been proposed to protect privacy by obfuscating or encrypting information. Fuzzy processing will limit matching accuracy, while encrypted information will increase the time cost of data computation. Therefore, this paper proposes a privacy-preserving map retrieval task assignment scheme (pMATE), which can divide the map and accurately retrieve the optimal workers according to this division. In pMATE, relevant information about tasks and workers is encrypted, and neighboring workers are searched based on the task presence partition. The task location can also be hidden in that partition. Partitioned retrieval reduces the amount of encrypted data needed to be matched. Furthermore, to reduce the problem of multiple communications during encrypted data comparison, we propose the Find MinNumber (FMN) algorithm, which can determine the optimal worker or top-k optimal workers need only two communications. Experimental evaluations of real-world data show that pMATE is efficient and accurate.