Task assignment optimization in knowledge-intensive crowdsourcing

摘要: We present SmartCrowd, a framework for optimizing task assignment in knowledge-intensive crowdsourcing (KI-C). SmartCrowd distinguishes itself by formulating, the first time, problem of worker-to-task KI-C as an optimization problem, proposing efficient adaptive algorithms to solve it and accounting human factors, such worker expertise, wage requirements, availability inside process. rigorous theoretical analyses propose optimal approximation with guarantees, which rely on index pre-computation maintenance. perform extensive performance quality experiments using real synthetic data demonstrate that approach is necessary achieve assignments high-quality under guaranteed cost budget.