[Colloquium] REMINDER: 4/11 Talks at TTIC: YangFeng Ji, Georgia Tech

[Mary Marre]

When:     Monday, April 11th at 11:00 am

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

Who:       YangFeng Ji, Georgia Tech


Title:
Distributed Representation Learning for Discourse Processing

Abstract:
Discourse processing is to identify coherent relations, such as contrast
and causal relation, from well-organized texts. These relations can benefit
both research and applications in natural language processing from various
aspects, such as recognizing the major opinion from a product review, or
generating words from context. Discourse relation identification requires
intensive semantic understanding of text, especially when no word (e.g.,
but) can signal the relations. Most prior work relies on sparse
representation constructed from surface features (including, word pairs,
POS tags, etc.), which fails to encode and share enough semantic
information. As an alternative, distributed representation uses dense
vectors, which are able to share information efficiently.

In this talk, I will present my work on learning distributed representation
for discourse processing. I propose a unified framework to learn both
distributed representation and discourse models of texts jointly. The
representation learned with discourse modeling facilitates the final
processing tasks. The evaluation shows that our systems outperform prior
work with only surface-form representations. Moreover, I will also discuss
how the idea of representation learning can be extended to other discourse
relevant tasks, for example, language generation with discourse
information. At the end of talk, I will demonstrate the reward of using
discourse information for document-level machine translation and sentiment
analysis.



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 <%28773%29%20834-1757>*
*f: (773) 357-6970 <%28773%29%20357-6970>*
*mmarre at ttic.edu <mmarre at ttic.edu>*
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.cs.uchicago.edu/pipermail/colloquium/attachments/20160410/12cf295c/attachment.htm