Avicaching: A Two Stage Game for Bias Reduction in Citizen Science

Citizen science projects have been very successful at collecting rich datasets for different applications. However, the data collected by the citizen scientists are often biased, more aligned with the citizens' preferences rather than scientific objectives. We introduce a novel two-stage game for reducing data-bias in citizen science in which the game organizer, a citizen-science program, incentivizes the agents, the citizen scientists, to visit under-sampled areas. We provide a novel way of encoding this two-stage game as a single optimization problem, cleverly folding (an approximation of) the agents' problems into the organizer's problem. We present several new algorithms to solve this optimization problem as well as a new structural SVM approach to learn the parameters that capture the agents' behaviors, under different incentive schemes. We apply our methodology to eBird, a well-established citizen-science program for collecting bird observations, as a game called Avicaching. We deployed Avicaching in two New York counties (March 2015), with a great response from the birding community, surpassing the expectations of the eBird organizers and bird-conservation experts. The field results show that the Avicaching incentives are remarkably effective at encouraging the bird watchers to explore under-sampled areas and hence alleviate the eBird's data bias problem.