Computer Science Colloquium - Learning Language Through Interaction

Friday, February 22 @ 2:35pm in Wege (TCL 123)

“Learning Language Through Interaction”

Artificial intelligence systems in general, and natural language processing systems in particular, which are build using machine learning techniques are amazingly effective when plentiful labeled training data exists for the task/domain of interest. Unfortunately, for broad coverage (both in task and domain and language) language understanding, we’re unlikely to ever have sufficient labeled data, and systems must find some other way to learn. I’ll describe work we’ve done building methods that can learn from interactions applied to two canonical NLP problems: machine translation and question answering. In the former, we develop techniques for collaborating with people; the latter, for competing with them. This talk highlights joint work with a number of wonderful students at collaborators at the University of Maryland, UC Boulder and Microsoft Research.

Hal Daumé III is a professor of Computer Science at the University of Maryland, College Park; he is currently on leave at Microsoft Research, New York City. He holds joint appointments in UMIACS and Linguistics. He was previously an assistant professor in the School of Computing at the University of Utah. His primary research interest is in developing new learning algorithms for prototypical problems that arise in the context of natural language processing and artificial intelligence, with a focus on bridging background domain knowledge with statistical learning. He associates himself most with conferences like ACL, ICML, NeurIPS and EMNLP, where he has published over 100 papers. He has received several “best of” awards, including at ACL 2018, NAACL 2016, NeurIPS 2015, CEAS 2011 and ECML 2009. He has been program chair for NAACL 2013 (and chair of its executive board), and will be program chair for ICML 2020; he was an inaugural diversity and inclusion co-chair at NeurIPS 2018. He earned his PhD at the University of Southern California with a thesis on structured prediction for language (his advisor was Daniel Marcu). He spent the summer of 2003 working with Eric Brill in the machine learning and applied statistics group at Microsoft Research. Prior to that, he studied math (mostly logic) at Carnegie Mellon University.