Statistical Machine Translation with Readability Constraints

Sara Stymne, Jörg Tiedemann, Christian Hardmeier, Joakim Nivre

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

This paper presents experiments with document-level machine translation with readability constraints. We describe the task of producing simplified translations from a given source with the aim to optimize machine translation for specific target users such as language learners. In our approach, we introduce global features that are known to affect readability into a document-level SMT decoding framework. We show that the decoder is capable of incorporating those features and that we can influence the readability of the output as measured by common metrics. This study presents the first attempt of jointly performing machine translation and text simplification, which is demonstrated through the case of translating parliamentary texts from English to Swedish.
Original languageEnglish
Title of host publicationProceedings of the 19th Nordic Conference of Computational Linguistics (NODALIDA 2013)
EditorsStephan Oepen, Kristin Hagen, Janne Bondi Johannessen
Place of PublicationOslo, Norway
PublisherLinköping University Electronic Press, Sweden
Pages375-386
Number of pages12
ISBN (Electronic)978-91-7519-589-6
Publication statusPublished - 24 May 2013
Event19th Nordic Conference of Computational Linguistics - Oslo, Norway
Duration: 22 May 201324 May 2013
https://www.hf.uio.no/iln/english/research/news-and-events/events/conferences/2013/nodalida/index.html

Publication series

Name
PublisherLinköping University Electronic Press
Volume85
ISSN (Electronic)1650-3740

Conference

Conference19th Nordic Conference of Computational Linguistics
Abbreviated titleNODALIDA 2013
CountryNorway
CityOslo
Period22/05/1324/05/13
Internet address

Keywords

  • Machine Translation
  • Text Simplification
  • Readability

Fingerprint Dive into the research topics of 'Statistical Machine Translation with Readability Constraints'. Together they form a unique fingerprint.

Cite this