Crowdsourcing Task Assignment with Online Profile Learning

In this paper, we present a reference framework called Argo+ for worker-centric crowdsourcing where task assignment is characterized by feature-based representation of both tasks and workers and learning techniques are exploited to online predict the most appropriate task to execute for a requesting worker. On the task side, features are used to represent requirements expressed in terms of knowledge expertise that are asked to workers for being involved in the task execution. On the worker side, features are used to compose a profile, namely a structured description of the worker capabilities in executing tasks. Experimental results obtained on a real crowdsourcing campaign are discussed by comparing the performance of Argo+ against a baseline with conventional task assignment techniques