On the Power of Bonus in Posted Pricing for Crowdsourcing
Suho Shin, Ho‐Yong Choi, Jungseul Ok, Yung Yi · 2018
In many practical crowdsourcing systems such as Amazon Mechanical Turk, posted pricing, where a pricing rule is published a priori, and workers then decide their task acceptance, is widely used due to its simplicity. With the goal of designing efficient posted pricing, which recruits more high-quality workers with less budget, we study the impact of the following two ideas in posted pricing: (i) personalization and (ii) giving bonus to more qualified task completion. In the Bayesian setting where only prior distribution of workers' profiles are available, we first study the Price of Agnosticity (PoA) that quantifies the utility gap between personalized and common pricing policies, where PoA is shown to be bounded within a constant factor under some mild conditions. Next, we analytically prove that the impact of bonus is significant in common pricing, which implies that a complex personalized pricing with privacy concerns can be replaced by a simple common pricing once it is equipped with a simple bonus payment. We validate our analytical findings through extensive real experiments in Amazon Mechanical Turk, one of the most popular crowdsourcing platforms.