Search Descriptions

Main Topics

Search Publications


author

title

other

year

Task Based Metrics

Task based evaluation of machine translation tests the usefulness of machine translation output to perform a given task.

Task Based Metrics is the main subject of 12 publications.

Publications

One test for the understanding of translated texts is to answer who, when, and where questions (Voss and Tate, 2006). Jones et al. (2006) use a foreign language learner test for measuring speech-to-speech machine translation performance. The adaptation of such a test (the Defense Language Proficiency Test, DLPT) for machine translation revealed that Arabic–English machine translation performance achieves a passing grade up to level 2+, but has a relatively weak performance on basic levels.

Benchmarks

Discussion

Related Topics

New Publications

  • Ageeva et al. (2015)
  • Sugiyama et al. (2015)
  • Russell and Gillespie (2016)
  • Matsuzaki et al. (2015)
  • Berka et al. (2011)
  • Wang and Boitet (2013)
  • Ogden et al. (2009)
  • Tate (2008)
  • Schneider et al. (2010)
  • Jones et al. (2007)

Actions

Download

Contribute