Improving Satisfaction in Crowdsourcing Platforms
Griffith Samore, Jonah Bates, Xiao Chen · 2021
Crowdsourcing platforms have gained popularity in recent years. They allow requesters to quickly find workers to complete small tasks and workers to undertake pieces of work for self-fulfillment. The current crowdsourcing models do the task assignment either from the requester's side (server assigned task mode) or the worker's side (worker selected task mode). The satisfaction of both sides is not fully considered. Furthermore, there is a lack of tools to help them make decisions based on complex information and their preferences. Therefore, in this paper, we propose a new crowdsourcing platform that takes the satisfaction of both the requesters and workers into account so as to improve the quality of the platform. We adopt Analytic Hierarchy Process (AHP) to automatically generate preference lists for both parties that best reflect their interests. We propose a stable matching (SM) algorithm to pair the workers and tasks according to their preference lists. Simulation results show that our platform has higher satisfaction scores than the existing ones and the one that uses random assignment.