Change search
ReferencesLink to record
Permanent link

Direct link
Modelling Pronominal Anaphora in Statistical Machine Translation
Fondazione Bruno Kessler, Human Language Technologies, Trento, Italy. (Datorlingvistik)
Fondazione Bruno Kessler, Trento. (Human Language Technologies Unit)
2010 (English)In: Proceedings of the 7th International Workshop on Spoken Language Translation, Paris. / [ed] Marcello Federico et al., Paris, 2010, 283-289 p.Conference paper (Refereed)
Abstract [en]

Current Statistical Machine Translation (SMT) systems translate texts sentence by sentence without considering any cross-sentential context. Assuming independence between sentences makes it difficult to take certain translation decisions when the necessary information cannot be determined locally. We argue for the necessity to include cross-sentence dependencies in SMT. As a case in point, we study the problem of pronominal anaphora translation by manually evaluating German-English SMT output. We then present a word dependency model for SMT, which can represent links between word pairs in the same or in different sentences. We use this model to integrate the output of a coreference resol- ution system into English-German SMT with a view to improving the translation of anaphoric pronouns.

Place, publisher, year, edition, pages
Paris, 2010. 283-289 p.
National Category
Language Technology (Computational Linguistics)
URN: urn:nbn:se:uu:diva-154490OAI: diva2:420761
IWSLT (International Workshop on Spoken Language Translation); Paris, France; December 2nd and 3rd, 2010.
Available from: 2011-06-07 Created: 2011-06-05 Last updated: 2013-09-18Bibliographically approved

Open Access in DiVA

hardmeier-federico:iwslt2010(363 kB)627 downloads
File information
File name FULLTEXT01.pdfFile size 363 kBChecksum SHA-512
Type fulltextMimetype application/pdf

Other links

Conference website

Search in DiVA

By author/editor
Hardmeier, Christian
Language Technology (Computational Linguistics)

Search outside of DiVA

GoogleGoogle Scholar
Total: 627 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

Total: 561 hits
ReferencesLink to record
Permanent link

Direct link