 |
Communication dans un congrès
Priming Neural Machine Translation
Abstract : Priming is a well known and studied psychology phenomenon based on the prior presentation of one stimulus (cue) to influence the processing of a response.
In this paper, we propose a framework to mimic the process of priming in the context of neural machine translation (NMT).
We evaluate the effect of using similar translations as priming cues on the NMT network.
We propose a method to inject priming cues into the NMT network
and compare our framework to other mechanisms that perform micro-adaptation during inference.
Overall, experiments conducted in a multi-domain setting confirm that adding priming cues in the NMT decoder can go a long way towards improving the translation accuracy.
Besides, we show the suitability of our framework to gather valuable information for an NMT network from monolingual resources.
Keywords :
translation memories
Machine translation
https://hal.archives-ouvertes.fr/hal-03013196
Contributeur : Limsi Publications <>
Soumis le : jeudi 19 novembre 2020 - 15:22:15
Dernière modification le : lundi 22 février 2021 - 16:21:22
Archivage à long terme le : : samedi 20 février 2021 - 19:04:32
Contributeur : Limsi Publications <>
Soumis le : jeudi 19 novembre 2020 - 15:22:15
Dernière modification le : lundi 22 février 2021 - 16:21:22
Archivage à long terme le : : samedi 20 février 2021 - 19:04:32
Fichier
2020.wmt-1.63.pdf
Fichiers éditeurs autorisés sur une archive ouverte