Market-Based Incentive Mechanism Design for Crowdsourcing
Feng Tian, Liling Huang · 2018
Participants coverage is an important role in the crowdsourcing-based applications. However most of the existing incentive mechanisms for the crowdsourcing has focused on how to allocate tasks to the participants to maximize the social welfare, and none of them consider the problem of participant coverage hole created by the uneven distribution of participants. In this paper, we propose a market-based incentive mechanism for crowdsourcing, where the platform motivates the participants to move to the coverage hole and complete the sensing tasks there from a market-based perspective. The market-based incentive mechanism is built on a novel reverse auction framework with reserve price, where the reserve price of a sensing task is the maximum payment for a participant to complete this task. The platform systematically reduces the reserve price of tasks in popular areas and participants are stimulated to complete the tasks in unpopular areas. Each round of reverse auction consists of a winning participant selection problem and a payment determination problem. Since the task allocation problem is NP-hard, we propose a greedy algorithm to solve it. We also design a critical payment policy to guarantee that participants declare their cost truthfully. Evaluation results show that the proposed mechanism outperforms existing solutions under various conditions.