[Colloquium] 1/17 Talks at TTIC: Sam Wiseman, Harvard University

[Mary Marre via Colloquium]

When:     Wednesday, January 17th at *10:30 am *

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

Who:       Sam Wiseman, Harvard University


Title:       Neural, Search-based Structured Prediction for Natural
Language Processing

Abstract: Much of the recent empirical success in natural language
processing has relied on the use of neural networks to compute expressive,
global representations of sentences and documents. It remains a challenge,
however, to incorporate such representations into systems that produce
structured outputs, such as natural language sentences, rather than
discrete labels. This challenge arises from the intractability of
performing inference in the presence of these non-local representations. In
this talk I will argue that search-based structured prediction, which
involves training a model to search incrementally for its predictions,
offers a particularly convenient and powerful approach for incorporating
neural representations into structured prediction systems. I will focus in
particular on techniques for training the ubiquitous sequence-to-sequence
model in this way, with application to tasks including parsing and machine
translation. I will also address techniques for tackling document-level
modeling in this framework, with application to coreference resolution, a
fundamental task in information extraction.


Host: Kevin Gimpel <kgimpel at ttic.edu>




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/20180110/d4f04cf3/attachment.html>