Task Assignments in Complex Collaborative Crowdsourcing

Abstract

At present, crowdsourcing platforms are mainly developed for Micro-tasks such as data annotation and image recognition. However, complex collaborative crowdsourcing tasks, which usually contain multiple interdependent subtasks, are emerging and require a variety of workers with different skills to cooperate with each other. To solve the worker-task matching problem in the complex collaborative crowdsourcing, we propose a complex crowdsourcing task assignment algorithm ($C^3TA$), which models the complex collaborative crowdsourcing task assignment problem as a combinatorial optimization problem based on the maximum flow and compute the optimal solution to task assignment with a Slide-Container Queue (SCP). The experimental results show that the algorithm can effectively assign complex collaborative crowdsourcing tasks with the constraint of multi-skill workers and sequential subtasks and maximize the number of assigned tasks.