Dynamic Pricing in Spatial Crowdsourcing: A Matching-Based Approach

Pricing is essential for the commercial success of spatial crowdsourcing applications. The spatial crowdsourcing platform prices tasks according to the demand and the supply in the crowdsourcing market to maximize its total revenue. Traditional pricing strategies seek a unified optimal price for a single global market. Yet spatial crowdsourcing needs to dynamically price for multiple local markets fragmented by the spatiotemporal distributions of tasks and workers and the mobility of workers. Dynamic pricing in spatial crowdsourcing is challenging because the supply in local markets can be limited and dependent, leading to global dependencies when pricing tasks in each local market. To this end, we define the Global Dynamic Pricing (GDP) problem in spatial crowdsourcing. We flirt her propose a MAtching-based Pricing Strategy (MAPS) with guaranteed bound, which efficiently approximates the expected total revenue for markets with limited supply, effectively distributes the dependent supply and dynamically prices the tasks. Extensive evaluations on both synthetic and real-world datasets demonstrate the effectiveness and efficiency of MAPS.