Amélioration de la traduction automatique d'un corpus annoté

Abstract : Improvement of the automatic translation of an annotated corpus In this article, we present a method to improve the automatic translation of an annotated corpus and transfer its annotations from English to any target language. The idea is to improve method of (Nasirud-din et al., 2015) which leads to many untranslated segments, duplications and disorders. We propose a pre-treatment process for the English SemCor, to adapt it to the statistical machine translation system, as well as a post-treatment process for the output of SMT. We show an increase of 2,9 points in terms of F1 score on a Word Sense Disambiguation task which proves the effectiveness of our method.
Document type :
Conference papers
Liste complète des métadonnées

Cited literature [6 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-01680553
Contributor : Marwa Hadj Salah <>
Submitted on : Tuesday, January 16, 2018 - 11:41:31 AM
Last modification on : Monday, February 11, 2019 - 4:36:02 PM
Document(s) archivé(s) le : Wednesday, May 23, 2018 - 5:54:43 PM

File

T63.pdf
Publisher files allowed on an open archive

Identifiers

  • HAL Id : hal-01680553, version 1

Collections

Citation

Marwa Hadj Salah, Hervé Blanchon, Mounir Zrigui, Didier Schwab. Amélioration de la traduction automatique d'un corpus annoté. JEP-TALN-RECITAL 2016, Jul 2016, Paris, France. ⟨hal-01680553⟩

Share

Metrics

Record views

191

Files downloads

63