The application of structured learning in natural language processing
The application of structured learning in natural language processing
We propose a structured learning approach, max-margin structure (MMS), which is targeted at natural language processing (NLP) tasks. The architecture of our approach is shown to capture structural aspects of the problem domains, leading to demonstrable performance improvements on two NLP tasks: part-of-speech tagging and statistical machine translation (SMT). We present a perceptron-based online learning algorithm to train the model and demonstrate desirable computational scaling behavior over traditional optimisation methods.
Ni, Yizhao
0452e056-90d0-4feb-a97b-ff2689b6b492
Saunders, Craig
26634635-4d4d-4469-b9ec-1d68788aa47a
Szedmak, Sandor
c6a84aa3-2956-4acf-8293-a1b676f6d7d8
Niranjan, Mahesan
5cbaeea8-7288-4b55-a89c-c43d212ddd4f
May 2010
Ni, Yizhao
0452e056-90d0-4feb-a97b-ff2689b6b492
Saunders, Craig
26634635-4d4d-4469-b9ec-1d68788aa47a
Szedmak, Sandor
c6a84aa3-2956-4acf-8293-a1b676f6d7d8
Niranjan, Mahesan
5cbaeea8-7288-4b55-a89c-c43d212ddd4f
Ni, Yizhao, Saunders, Craig, Szedmak, Sandor and Niranjan, Mahesan
(2010)
The application of structured learning in natural language processing.
Machine Translation.
Abstract
We propose a structured learning approach, max-margin structure (MMS), which is targeted at natural language processing (NLP) tasks. The architecture of our approach is shown to capture structural aspects of the problem domains, leading to demonstrable performance improvements on two NLP tasks: part-of-speech tagging and statistical machine translation (SMT). We present a perceptron-based online learning algorithm to train the model and demonstrate desirable computational scaling behavior over traditional optimisation methods.
Text
NiMachineTranslation2010.pdf
- Accepted Manuscript
Restricted to Registered users only
Request a copy
More information
Published date: May 2010
Organisations:
Southampton Wireless Group
Identifiers
Local EPrints ID: 271262
URI: http://eprints.soton.ac.uk/id/eprint/271262
ISSN: 0922-6567
PURE UUID: cd3254a3-fd03-429f-9d8a-257e05212a15
Catalogue record
Date deposited: 14 Jun 2010 12:43
Last modified: 15 Mar 2024 03:29
Export record
Contributors
Author:
Yizhao Ni
Author:
Craig Saunders
Author:
Sandor Szedmak
Author:
Mahesan Niranjan
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics