CobaltF: a fluent metric for MT evaluation

  • Authors
  • Fomicheva M, Bel N, Specia L, da Cunha I, Malinovskiy A
  • UPF authors
  • BEL RAFECAS, NÚRIA; DA CUNHA FANEGO, IRIA;
  • Authors of the book
  • -
  • Book title
  • The 54th Annual Meeting of the Association for Computational Linguistics : Proceedings of the Conference, Vol. 2
  • Publisher
  • Association for Computational Linguistics
  • Publication year
  • 2016
  • Pages
  • 483-490
  • ISBN
  • 978-1-945626-01-2
  • Abstract
  • The vast majority of Machine Translation (MT) evaluation approaches are based on the idea that the closer the MT output is to a human reference translation, the higher its quality. While translation quality has two important aspects, adequacy and fluency, the existing referencebased metrics are largely focused on the former. In this work we combine our metric UPF-Cobalt, originally presented at the WMT15 Metrics Task, with a number of features intended to capture translation fluency. Experiments show that the integration of fluency-oriented features significantly improves the results, rivalling the best-performing evaluation metrics on the WMT15 data.
  • Complete citation
  • Fomicheva M, Bel N, Specia L, da Cunha I, Malinovskiy A. CobaltF: a fluent metric for MT evaluation. In: -. The 54th Annual Meeting of the Association for Computational Linguistics Proceedings of the Conference, Vol. 2. 1 ed. East Stroudsburg PA: Association for Computational Linguistics; 2016. p. 483-490.