Combined Crowdsourcing Task Auction Mechanism Based on Stable Matching.

Lanyu Zhang,Mingjun Xiao,Hui Zhao,JianChun Liu

DOI: https://doi.org/10.1109/hpcc/smartcity/dss.2019.00263

2019-01-01

Abstract:Crowdsourcing allows requester to allocate tasks to a group of workers on the internet to make use of their collective intelligence. In this paper, we consider the complementarity among different types of tasks, which means the cost of a worker to complete a task combination is less than the sum of the costs that the worker completes each task in the combination separately. In the crowdsourcing, task assignment is one of the most essential problems, we formulate the crowdsourcing task assignment as a many-to-many matching problem. However, most existing works ignore the preferences of crowdsourcers and workers. Since both of workers and crowdsourcers are selfish individuals who are not necessarily aligned with the system optimization. They can violate the task assignment to obtain a higher utility. Different from the traditional task assignment mechanisms which focus on social welfare maximization, we not only consider the preferences of crowdsourcers and workers, but also obtain a stable matching result for the many-to-many matching problem. Payment determination is also a essential problem in the crowdsourcing system, so we compute the corresponding payment profiles for both crowdsourcers and workers. Extensive simulations have been performed, and the result demonstrates that our proposed mechanism not only obtains a stable matching but also has a high individual utility for each worker and each crowdsourcer. Workers and crowdsourcers would comply with the allocation results.