eprintid: 4604
rev_number: 6
eprint_status: archive
userid: 404
dir: disk0/00/00/46/04
datestamp: 2021-08-15 08:42:28
lastmod: 2021-08-15 08:42:28
status_changed: 2021-08-15 08:42:28
type: article
metadata_visibility: show
creators_name: Nguyen, Minh Thuan
creators_name: Nguyen, Phuong Thai
creators_name: Nguyen, Van Vinh
creators_name: Nguyen Hoang, Minh Cong
creators_id: thuannm@vnu.edu.vn
creators_id: thainp@vnu.edu.vn
creators_id: vinhnv@vnu.edu.vn
creators_id: minhcongnguyen1508@gmail.com
title: Iterative Multilingual Neural Machine Translation for Less-Common and Zero-Resource Language Pairs
ispublished: pub
subjects: IT
divisions: fac_fit
abstract: Research on providing machine translation systems for unseen language pairs is gaining increasing attention in recent years. However, the quality of their systems is poor for most language pairs, especially for less-common pairs such as Khmer-Vietnamese. In this paper, we show a simple iterative traininggenerating-filtering-training process that utilizes all available pivot parallel data to generate synthetic data for unseen directions. In addition, we propose a filtering method based on word alignments and the longest parallel phrase to filter out noise sentence pairs in the synthetic data. Experiment results on zero-shot Khmer→Vietnamese and Indonesian→Vietnamese directions show that our proposed model outperforms some strong baselines and achieves a promising result under the zero-resource condition on ALT benchmarks. Besides, the results also indicate that our model can easily improve their quality with a small amount of real parallel data.
date: 2020-10
date_type: published
publisher: Association for Computational Linguistics
official_url: https://aclanthology.org/2020.paclic-1.24
id_number: 2020.paclic-1.24
full_text_status: public
publication: Proceedings of the 34th Pacific Asia Conference on Language, Information and Computation
pagerange: 207-215
refereed: FALSE
citation:   Nguyen, Minh Thuan and Nguyen, Phuong Thai and Nguyen, Van Vinh and Nguyen Hoang, Minh Cong  (2020) Iterative Multilingual Neural Machine Translation for Less-Common and Zero-Resource Language Pairs.  Proceedings of the 34th Pacific Asia Conference on Language, Information and Computation .   pp. 207-215.      
document_url: https://eprints.uet.vnu.edu.vn/eprints/id/eprint/4604/1/2020.paclic-1.24.pdf