Machine Translation

, Volume 27, Issue 3–4, pp 213–238

Quality estimation for machine translation: some lessons learned

  • Guillaume Wisniewski
  • Anil Kumar Singh
  • François Yvon
Article

DOI: 10.1007/s10590-013-9141-9

Cite this article as:
Wisniewski, G., Singh, A.K. & Yvon, F. Machine Translation (2013) 27: 213. doi:10.1007/s10590-013-9141-9

Abstract

The dissemination of statistical machine translation (SMT) systems in the professional translation industry is still limited by the lack of reliability of SMT outputs, the quality of which varies to a great extent. A critical piece of information would be for MT systems to automatically assess their output translations with automatically derived quality measures. Predicting quality measures was indeed the goal of a shared task at the Workshop on SMT in 2012. In this contribution, we first report our results for this shared task, detailing the features that we found to be the most predictive of quality. In the latter part, we reexamine the shared task data and protocol and show that several factors actually contributed to the difficulty of the task, and discuss alternative evaluation designs.

Copyright information

© Springer Science+Business Media Dordrecht 2013

Authors and Affiliations

  • Guillaume Wisniewski
    • 1
  • Anil Kumar Singh
    • 2
  • François Yvon
    • 1
  1. 1.LIMSI—Université Paris SudOrsayFrance
  2. 2.LIMSI—CNRS OrsayFrance

Personalised recommendations