Scopus: 0 citations, Google Scholar: citations
Comparing NMT and PBSMT for post-editing in-domain formal texts : a case study
Álvarez Vidal, Sergi (Universitat Pompeu Fabra)
Badia, Toni (Universitat Pompeu Fabra)
Oliver, Antoni (Oliver Gonzàlez) (Universitat Oberta de Catalunya)

Imprint: Berlin (Alemanya) : Language Science Press, 2021
Description: 14 pàg.
Abstract: This paper details a comparative analysis between phrase-based statistical machine translation (PBSMT) and neural machine translation (NMT) for English-Spanish in-domain medical documents using human rankings, fluency and adequacy, and post-editing (technical and temporal) effort, performed by professional translators. When MT output is ranked against translations performed by professional translators, results show a clear preference for human translations, with NMT in the second position. Regarding MT outputs, NMT is perceived as more fluent and conveying better the meaning of the source sentence. Despite this preference, postediting temporal effort does not improve significantly in NMT compared to PBSMT, although technical effort is reduced.
Rights: Aquest document està subjecte a una llicència d'ús Creative Commons. Es permet la reproducció total o parcial, la distribució, la comunicació pública de l'obra i la creació d'obres derivades, fins i tot amb finalitats comercials, sempre i quan es reconegui l'autoria de l'obra original. Creative Commons
Language: Anglès
Document: Capítol de llibre ; recerca ; Versió publicada
Published in: Translation, interpreting, cognition. The way out of the box, 2021, p. 33-46, ISBN 9783985540006

DOI: 10.5281/zenodo.4545031


14 p, 122.9 KB

The record appears in these collections:
Research literature > UAB research groups literature > Research Centres and Groups (research output) > Arts and Humanities > Grup d'estudi de la literacitat en l’ensenyament i l’aprenentatge de segones llengües i traducció (GELEA2LT)
Books and collections > Book chapters

 Record created 2024-10-31, last modified 2026-01-17



   Favorit i Compartir