If anyone is interested in working on language and translation, this could be an interesting project with large general benefit! -Matt J.
---------- Forwarded message --------- From: Emily M. Bender <eben...@uw.edu> Date: Thu, Dec 13, 2018 at 18:36 Subject: [dub] Design project idea To: <d...@dub.washington.edu> Dear all, I have an idea I'd like to pursue which I think requires collaboration between someone with expertise in design (especially UI design) and a computational linguist (like me!). This is just a quick sketch of the idea, because I'd like to know if this appeals to anyone on this list. Basically, the problem I'd like to tackle is the way that certain NLP technology makes opaque both the source of its information and its level of uncertainty. I think this is a contributing factor to problems like this one: https://www.theguardian.com/technology/2017/oct/24/facebook-palestine-israel-translates-good-morning-attack-them-arrest I think one possible (partial) solution here is to find a way to make that information visible to naive users. I know that when I approach machine translation systems, for example, I bring a lot of skepticism about the results based on what I know about how the systems work and how translation works in general. What I'd like to explore is whether it's possible to expose information about the system in the right way to engender similar skepticism in non-linguist users. Does this pique anyone's interest? Emily -- Emily M. Bender Professor, Department of Linguistics University of Washington Twitter: @emilymbender _______________________________________________ dub mailing list d...@dub.washington.edu https://dubber.cs.washington.edu/mailman/listinfo/dub
_______________________________________________ change mailing list change@change.washington.edu https://changemm.cs.washington.edu/mailman/listinfo/change