Document-level machine translation as a re-translation process

Eva Martínez Garcia, Cristina España-Bonet, Lluis Marques

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Most of the current Machine Translation systems are designed to translate a document sentence by sentence ignoring discourse information and producing incoherencies in the final translations. In this paper we present some documentlevel- oriented post-processes to improve translations' coherence and consistency. Incoherences are detected and new partial translations are proposed. The work focuses on studying two phenomena: words with inconsistent translations throughout a text and also, gender and number agreement among words. Since we deal with specific phenomena, an automatic evaluation does not re ect significant variations in the translations. However, improvements are observed through a manual evaluation.

Original languageEnglish
Pages (from-to)103-110
Number of pages8
JournalProcesamiento de Lenguaje Natural
Volume53
Publication statusPublished - 1 Jan 2014

Fingerprint

evaluation
Machine Translation
Translation Process
discourse
gender
Evaluation
Post-process
Gender Agreement
Incoherence
Number Agreement
Discourse
Machine Translation System
coherence

Keywords

  • Coherence
  • Coreference
  • Discourse
  • Statistical machine translation

ASJC Scopus subject areas

  • Language and Linguistics
  • Linguistics and Language
  • Computer Science Applications

Cite this

Document-level machine translation as a re-translation process. / Garcia, Eva Martínez; España-Bonet, Cristina; Marques, Lluis.

In: Procesamiento de Lenguaje Natural, Vol. 53, 01.01.2014, p. 103-110.

Research output: Contribution to journalArticle

Garcia, Eva Martínez ; España-Bonet, Cristina ; Marques, Lluis. / Document-level machine translation as a re-translation process. In: Procesamiento de Lenguaje Natural. 2014 ; Vol. 53. pp. 103-110.
@article{4a44acb5f61e42d9bc17b4ca15f52c5c,
title = "Document-level machine translation as a re-translation process",
abstract = "Most of the current Machine Translation systems are designed to translate a document sentence by sentence ignoring discourse information and producing incoherencies in the final translations. In this paper we present some documentlevel- oriented post-processes to improve translations' coherence and consistency. Incoherences are detected and new partial translations are proposed. The work focuses on studying two phenomena: words with inconsistent translations throughout a text and also, gender and number agreement among words. Since we deal with specific phenomena, an automatic evaluation does not re ect significant variations in the translations. However, improvements are observed through a manual evaluation.",
keywords = "Coherence, Coreference, Discourse, Statistical machine translation",
author = "Garcia, {Eva Mart{\'i}nez} and Cristina Espa{\~n}a-Bonet and Lluis Marques",
year = "2014",
month = "1",
day = "1",
language = "English",
volume = "53",
pages = "103--110",
journal = "Procesamiento de Lenguaje Natural",
issn = "1135-5948",
publisher = "Sociedad Espanola para el Procesamiento del Lenguaje Natural",

}

TY - JOUR

T1 - Document-level machine translation as a re-translation process

AU - Garcia, Eva Martínez

AU - España-Bonet, Cristina

AU - Marques, Lluis

PY - 2014/1/1

Y1 - 2014/1/1

N2 - Most of the current Machine Translation systems are designed to translate a document sentence by sentence ignoring discourse information and producing incoherencies in the final translations. In this paper we present some documentlevel- oriented post-processes to improve translations' coherence and consistency. Incoherences are detected and new partial translations are proposed. The work focuses on studying two phenomena: words with inconsistent translations throughout a text and also, gender and number agreement among words. Since we deal with specific phenomena, an automatic evaluation does not re ect significant variations in the translations. However, improvements are observed through a manual evaluation.

AB - Most of the current Machine Translation systems are designed to translate a document sentence by sentence ignoring discourse information and producing incoherencies in the final translations. In this paper we present some documentlevel- oriented post-processes to improve translations' coherence and consistency. Incoherences are detected and new partial translations are proposed. The work focuses on studying two phenomena: words with inconsistent translations throughout a text and also, gender and number agreement among words. Since we deal with specific phenomena, an automatic evaluation does not re ect significant variations in the translations. However, improvements are observed through a manual evaluation.

KW - Coherence

KW - Coreference

KW - Discourse

KW - Statistical machine translation

UR - http://www.scopus.com/inward/record.url?scp=84906858747&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84906858747&partnerID=8YFLogxK

M3 - Article

VL - 53

SP - 103

EP - 110

JO - Procesamiento de Lenguaje Natural

JF - Procesamiento de Lenguaje Natural

SN - 1135-5948

ER -