Skip to main navigation Skip to search Skip to main content

Stability vs. effectiveness: Improved sentence-level combination of machine translation based on weighted MBR

  • Harbin Institute of Technology

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

We describe an improved strategy to combine the outputs of machine translation on sentence-level balancing the stability and the effectiveness of the combination. The new method alternates the classical MBR-based sentence-level combination with weighted Minimum Bayes Risk (wMBR). During the calculation of the risk, we weight the hypotheses with the performance of the MT system, which is measured by the automatic evaluation metrics on the development data. In experiments, the wMBR-based method stably achieve better results than other sentence-level methods and get the best position in CWMT08 evaluation track outperforming the other word-level and sentence-level combination systems.

Original languageEnglish
Title of host publication2009 International Conference on Asian Language Processing
Subtitle of host publicationRecent Advances in Asian Language Processing, IALP 2009
Pages39-42
Number of pages4
DOIs
StatePublished - 2009
Event2009 International Conference on Asian Language Processing: Recent Advances in Asian Language Processing, IALP 2009 - Singapore, Singapore
Duration: 7 Dec 20099 Dec 2009

Publication series

Name2009 International Conference on Asian Language Processing: Recent Advances in Asian Language Processing, IALP 2009

Conference

Conference2009 International Conference on Asian Language Processing: Recent Advances in Asian Language Processing, IALP 2009
Country/TerritorySingapore
CitySingapore
Period7/12/099/12/09

Keywords

  • Combination
  • Machine translation
  • Minimum bayes risk

Fingerprint

Dive into the research topics of 'Stability vs. effectiveness: Improved sentence-level combination of machine translation based on weighted MBR'. Together they form a unique fingerprint.

Cite this