Word Representations in Factored Neural Machine Translation

Franck Burlot (1) , Mercedes Garcia-Martinez (2) , Loïc Barrault (2) , Fethi Bougares (2) , François Yvon (1)

1 LIMSI - Laboratoire d'Informatique pour la Mécanique et les Sciences de l'Ingénieur
2 LIUM - Laboratoire d'Informatique de l'Université du Mans

Translation into a morphologically rich language requires a large output vocabulary to model various morphological phenomena, which is a challenge for neural machine translation architectures. To address this issue, the present paper investigates the impact of having two output factors with a system able to generate separately two distinct representations of the target words. Within this framework, we investigate several word representations that correspond to different distributions of morpho-syntactic information across both factors. We report experiments for translation from English into two morphologically rich languages, Czech and Latvian, and show the importance of explicitly modeling target morphology.

Mots clés: machine translation; morphology; factored machine translation

Domaines: Informatique [cs]; Informatique et langage [cs.CL]

Franck Burlot, Mercedes Garcia-Martinez, Loïc Barrault, Fethi Bougares, François Yvon. Word Representations in Factored Neural Machine Translation. Conference on Machine Translation, Association for Computational Linguistics, Sep 2017, Copenhagen, Denmark. pp.43 - 55. ⟨hal-01618384⟩