Maximizing the Total Utility of Requesters in Crowdsourcing

Crowdsourcing coordinates a large group of workers online to do small tasks published by requesters on a crowd-sourcing platform. In the literature, the model used by many papers assumes that one task is given to and completed by one worker only. In this paper, we consider a model that extends this model in space and time. Based on our model, we formulate an optimization problem from the perspective of the requesters that maximizes the utility of all the requesters subject to the constraint that the total workload given to a worker should not exceed his capability. We then provide a solution to the problem and design distributed algorithms for the requesters and the workers to interact with each other in multiple rounds. After that, we give a concrete example to explain the solution and the algorithms. Then, we discuss the convergence speed of our algorithms using different methods. Finally, we conduct simulations to compare these convergence methods and draw conclusions.