Composed, Distributed Reflections on Semantics and Statistical Machine Translation

Add to your list(s) Download to your calendar using vCal

• Tim Baldwin, The University of Melbourne
• Friday 28 November 2014, 12:00-13:00
• FW26, Computer Laboratory.

If you have a question about this talk, please contact Tamara Polajnar.

Instances of successful application of explicit lexical and compositional semantic analysis to statistical machine translation (SMT) have been few and far between. Recent work on incorporating ideas from deep learning into SMT have lead to consistently impressive gains in translation accuracy, however, and “semantics” is now the talk of the town in NLP . In this talk, I will reflect on the rare instances of successful incorporation of lexical semantics in MT research, and the past failures of SMT at incorporating semantics. I will then go on to suggest different ways in which distributed semantic approaches can be incorporated into MT research, as well as (wildly) speculate on fresh directions for distributed semantic research that I believe will have even greater impact on MT.

This talk is part of the NLIP Seminar Series series.

This talk is included in these lists:

• All Talks (aka the CURE list)
• Cambridge Centre for Data-Driven Discovery (C2D3)
• Cambridge Forum of Science and Humanities
• Cambridge Language Sciences
• Cambridge talks
• Chris Davis' list
• Department of Computer Science and Technology talks and seminars
• FW26, Computer Laboratory
• Graduate-Seminars
• Guy Emerson's list
• Interested Talks
• Language Sciences for Graduate Students
• NLIP Seminar Series
• PMRFPS's
• Simon Baker's List
• Trust & Technology Initiative - interesting events
• bld31
• ndk22's list
• ob366-ai4er
• rp587

Note that ex-directory lists are not shown.