Abstract

In this talk I’ll describe recent progress in interactive neural machine translation, which is a human-in-the-loop machine learning task. First, I’ll motivate interactive MT, which is an idea that dates to the 1960s that has only recently been realized in commercially available systems. Next, I’ll describe recent research results for domain adaptation, inference, and word alignment. Finally, I’ll offer some brief reflections on doing research in a startup setting.

Biography

Spence Green is CEO of Lilt, which builds intelligent software to automate enterprise language translation. He received a PhD in computer science from Stanford University where he worked with Chris Manning and Jeff Heer. He also holds an M.S. from Stanford and a B.S. in computer engineering from the University of Virginia.