Privacy-preserving Task Recommendation for Cloud-based Crowdsourcing

Description

Crowdsourcing is an emerging technology that enables task-requesters (calledrequesters) to outsource tasks to a large group of undefined people (called workers). Asthe ever increasing number of tasks that are call-for-workers, workers can no longermanually select their interested tasks. A public server (called broker) is introduced tofacilitate the auto-matching of tasks with their potential workers. The broker is usuallydeployed in a public cloud due to the intensive computing and storage resources it needs.Information security and privacy becomes an important concern in this publiccrowdsourcing platform. Task requesters do not want to leak out any information abouttasks to the broker, and the workers also do not want reveal any information about theirpersonal interests to the broker. The primary goal of this project is to develop a Privacy-Preserving Task Recommendation (PPTR) scheme that can protect the privacy of bothrequesters and workers against the broker, while still enables the broker to performmatching between task requirements and worker interests. The project consists of threetasks: 1) Design a PPTR scheme that supports Multi-requesters and Multi-workers(M2M) crowdsourcing with keyword equality matching. The crowdsourcing systemsshould be able to support multiple requesters to publish tasks and multiple workers tosubscribe crowdsourced tasks. 2) Design a PPTR scheme that supports efficient numericrange matching for M2M crowdsourcing. This task aims to develop a PPTR scheme thatsupports task matching where task requirements may contain ranges of numeric values.3) Design a PPTR scheme that supports complex structures of range and keywordmatching for M2M crowdsourcing.?