I’m heading West for a June 10 meetup about the Mechanical Turk, hosted by the Turk experts, Dolores Labs:
The page contains a description of the program, and there’s space for a few more; just RSVP at the meetup page above.
Lukas (our host) just made up a title for my presentation (which is of joint work with Breck Baldwin and Emily Jamison), “Failure study of [a] biology task”. I’m going to be talking about some of the NLP tasks we’ve run successfully (named entity tagging and morphological analysis) and also one (linking MEDLINE to Entrez-Gene) that didn’t work. I’ll discuss some of the UI issues along the way. I’ll also describe the high level view of inferring gold standards and annotator accuracy, particularly with respect to its effect on applications. But don’t worry — I won’t be assuming the audience is full of Bayesian statistician.