Towards task recommendation in micro-task markets
Vamshi Ambati, Stephan Vogel, Jaime Carbonell · 2011
As researchers embrace micro-task markets for elicit-ing human input, the nature of the posted tasks moves from those requiring simple mechanical labor to requir-ing specific cognitive skills. On the other hand, increase is seen in the number of such tasks and the user popula-tion in micro-task market places requiring better search interfaces for productive user participation. In this pa-per we posit that understanding user skill sets and pre-senting them with suitable tasks not only maximizes the over quality of the output, but also attempts to maxi-mize the benefit to the user in terms of more success-fully completed tasks. We also implement a recommen-dation engine for suggesting tasks to users based on im-plicit modeling of skills and interests. We present results from a preliminary evaluation of our system using pub-licly available data gathered from a variety of human computation experiments recently conducted on Ama-zon’s Mechanical Turk.