Crowd-sourcing with uncertain quality - an auction approach

This article addresses two important issues in crowd-sourcing: ex ante uncertainty about the quality and cost of different workers and strategic behaviour. We present a novel multi-dimensional auction that incentivises the workers to make partial enquiry into the task and to honestly report quality-cost estimates based on which the crowd-sourcer can choose the worker that offers the best value for money. The mechanism extends second score auction design to settings where the quality is uncertain and it provides incentives to both collect information and deliver desired qualities.