[Colloquium] 11/29 TTIC Colloquium: Regina Barzilay, MIT

[Mary Marre via Colloquium]

*PLEASE NOTE: Special Day and Time!*

When:     Wednesday, November 29th at 3:00 p.m.

Where:    TTIC, 6045 S. Kenwood Avenue, 5th Floor, Room 526

Who:       Regina Barzilay, MIT

Title:       Moving Away from Big Data

Abstract:  In the last decade, the standard NLP datasets have increased in
size by orders of magnitude. For example, in textual entailment, the size
of the benchmark dataset changed from 7K instances in 2011  to 570K in
2015.  However, this dependence on large-scale annotations is not suitable
for many practical settings. This is especially true in medical
applications where privacy regulations prevent us from using crowdsourcing
to collect annotations. In this talk, I will describe our work on
developing information extraction systems that re designed for the
annotation scarce environment.  In particular, I will focus on multi-aspect
transfer models that reuse annotations across tasks, reinforcement learning
models that learn to find additional information sources to resolve
extraction ambiguity, and interpretable neural models that provide
rationales underlying their predictions.


Host: David McAllester <mcallester at ttic.edu>


For more information on the colloquium series or to subscribe to the
mailing list, please see http://www.ttic.edu/colloquium.php



Mary C. Marre
Administrative Assistant
*Toyota Technological Institute*
*6045 S. Kenwood Avenue*
*Room 504*
*Chicago, IL  60637*
*p:(773) 834-1757*
*f: (773) 357-6970*
*mmarre at ttic.edu <mmarre at ttic.edu>*
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.cs.uchicago.edu/pipermail/colloquium/attachments/20171124/12d29650/attachment.html>