Time: 4:00 - 5:00pm Venue: BR 3.02 Bancroft Road Teaching Rooms Peter Landin Building London E1 4NS
The Bayesian Crowd: scalable information combination for Citizen Science and Crowdsourcing
Abstract In Citizen Science and Crowdsourcing applications information from large numbers of agents need to be combined in an intelligent manner. For realistic deployment methods for such combination should conform to optimality wherever possible, yet scale well with large numbers of information sources and amounts of data. This talk will focus on Bayesian information aggregation models; we discuss how the use of approximate inference, based on variational learning, allows excellent scaling properties whilst retaining high performance. We showcase the breadth of applicability of the approach with examples from large Citizen Science and Crowdsourcing domains, feedback and user-task allocation mechanisms.