Please use this identifier to cite or link to this item: http://cris.unibe.edu.do/handle/123456789/320
Title: Improving a Japanese-Spanish machine translation system using Wikipedia medical articles
Autores: Ramírez Vidal, Jessica C.
Matsumoto, Y.
Muñoz N., Darwin C.
Researchers (UNIBE): Ramírez Vidal, Jessica C. 
Muñoz N., Darwin C. 
Affiliations: Facultad de Ciencias Básicas y Tecnológicas 
Facultad de Ciencias Básicas y Tecnológicas 
Research area: Tecnología
Keywords: Comparable Corpora; Dictionary; Ontology; Machine Translation
Issue Date: 2015
Publisher: Academy & Industry Research Collaboration Center (AIRCC)
Source: En: D. Wyld y J. Zizka (Eds.) Computer Science & Information Technology. Fifth International conference on Computer Science and Information Technology (CCSIT - 2015), pp. 111–116.
Project: Sistema de traducción automática japonés-español usando inglés como lengua intermedia 
Related Publication(s): Computer Science & Information Technology. Fifth International Conference on Computer Science and Information Technology (CCSIT - 2015)
Start page: 111
End page: 116
Conference: Fifth International Conference on Computer Science and Information Technology (CCSIT - 2015). Sydney, Australia, February 21-22, 2015
Abstract: 
The quality, length and coverage of a parallel corpus are fundamental features in the performance of a Statistical Machine Translation System (SMT). For some pair of languages there is a considerable lack of resources suitable for Natural Language Processing tasks. This paper introduces a technique for extracting medical information from the Wikipedia page. Using a medical ontological dictionary and then we evaluate on a Japanese-Spanish SMT system. The study shows an increment in the BLEU score.
URI: http://cris.unibe.edu.do/handle/123456789/320
DOI: 10.5121/csit.2015.50411
Appears in Collections:Publicaciones de la Facultad de Ciencias Básicas y Tecnológicas

Files in This Item:
File Description SizeFormat
csit53511.pdfFull text [open access]534.17 kBAdobe PDFView/Open
Show full item record Recommend this item

Google ScholarTM

Citations

Altmetric

Mentions

Dimensions

Citations


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.