Revisiting the Binary Linearization Technique for Surface Realization - Archive ouverte HAL Accéder directement au contenu
Article Dans Une Revue Proceedings of The 12th International Conference on Natural Language Generation Année : 2019

Revisiting the Binary Linearization Technique for Surface Realization

Résumé

End-to-end neural approaches have achieved state-of-the-art performance in many natural language processing (NLP) tasks. Yet, they often lack transparency of the underlying decision-making process, hindering error analysis and certain model improvements. In this work, we revisit the binary linearization approach to surface realization, which exhibits more interpretable behavior, but was falling short in terms of prediction accuracy. We show how enriching the training data to better capture word order constraints almost doubles the performance of the system. We further demonstrate that encoding both local and global prediction contexts yields another considerable performance boost. With the proposed modifications , the system which ranked low in the latest shared task on multilingual surface realization now achieves best results in five out of ten languages, while being on par with the state-of-the-art approaches in others.
Fichier principal
Vignette du fichier
W19-8635-yevgenyi.pdf (373.66 Ko) Télécharger le fichier
Origine : Fichiers éditeurs autorisés sur une archive ouverte
Loading...

Dates et versions

hal-02460309 , version 1 (31-01-2020)

Identifiants

  • HAL Id : hal-02460309 , version 1

Citer

Yevgenyi Puzikov, Claire Gardent, Ido Dagan, Iryna Gurevych. Revisiting the Binary Linearization Technique for Surface Realization. Proceedings of The 12th International Conference on Natural Language Generation, 2019, pp.268 - 278. ⟨hal-02460309⟩
30 Consultations
51 Téléchargements

Partager

Gmail Facebook X LinkedIn More