Machine Translation in Low-Resource Languages by an Adversarial Neural Network

Existing Sequence-to-Sequence (Seq2Seq) Neural Machine Translation (NMT) shows strong capability with High-Resource Languages (HRLs). However, this approach poses serious challenges when processing Low-Resource Languages (LRLs), because the model expression is limited by the training scale of parall...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Mengtao Sun, Hao Wang, Mark Pasquine, Ibrahim A. Hameed
Formato:	article
Lenguaje:	EN
Publicado:	MDPI AG 2021
Materias:	machine learning adversarial machine learning imbalanced datasets transfer learning Technology T Engineering (General). Civil engineering (General) TA1-2040 Biology (General) QH301-705.5 Physics QC1-999 Chemistry QD1-999
Acceso en línea:	https://doaj.org/article/c042e7e12b1748c3830acce4f976d5e0
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:c042e7e12b1748c3830acce4f976d5e0
record_format	dspace
spelling	oai:doaj.org-article:c042e7e12b1748c3830acce4f976d5e02021-11-25T16:39:30ZMachine Translation in Low-Resource Languages by an Adversarial Neural Network10.3390/app1122108602076-3417https://doaj.org/article/c042e7e12b1748c3830acce4f976d5e02021-11-01T00:00:00Zhttps://www.mdpi.com/2076-3417/11/22/10860https://doaj.org/toc/2076-3417Existing Sequence-to-Sequence (Seq2Seq) Neural Machine Translation (NMT) shows strong capability with High-Resource Languages (HRLs). However, this approach poses serious challenges when processing Low-Resource Languages (LRLs), because the model expression is limited by the training scale of parallel sentence pairs. This study utilizes adversary and transfer learning techniques to mitigate the lack of sentence pairs in LRL corpora. We propose a new Low resource, Adversarial, Cross-lingual (LAC) model for NMT. In terms of the adversary technique, LAC model consists of a generator and discriminator. The generator is a Seq2Seq model that produces the translations from source to target languages, while the discriminator measures the gap between machine and human translations. In addition, we introduce transfer learning on LAC model to help capture the features in rare resources because some languages share the same subject-verb-object grammatical structure. Rather than using the entire pretrained LAC model, we separately utilize the pretrained generator and discriminator. The pretrained discriminator exhibited better performance in all experiments. Experimental results demonstrate that the LAC model achieves higher Bilingual Evaluation Understudy (BLEU) scores and has good potential to augment LRL translations.Mengtao SunHao WangMark PasquineIbrahim A. HameedMDPI AGarticlemachine learningadversarial machine learningimbalanced datasetstransfer learningTechnologyTEngineering (General). Civil engineering (General)TA1-2040Biology (General)QH301-705.5PhysicsQC1-999ChemistryQD1-999ENApplied Sciences, Vol 11, Iss 10860, p 10860 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	machine learning adversarial machine learning imbalanced datasets transfer learning Technology T Engineering (General). Civil engineering (General) TA1-2040 Biology (General) QH301-705.5 Physics QC1-999 Chemistry QD1-999
spellingShingle	machine learning adversarial machine learning imbalanced datasets transfer learning Technology T Engineering (General). Civil engineering (General) TA1-2040 Biology (General) QH301-705.5 Physics QC1-999 Chemistry QD1-999 Mengtao Sun Hao Wang Mark Pasquine Ibrahim A. Hameed Machine Translation in Low-Resource Languages by an Adversarial Neural Network
description	Existing Sequence-to-Sequence (Seq2Seq) Neural Machine Translation (NMT) shows strong capability with High-Resource Languages (HRLs). However, this approach poses serious challenges when processing Low-Resource Languages (LRLs), because the model expression is limited by the training scale of parallel sentence pairs. This study utilizes adversary and transfer learning techniques to mitigate the lack of sentence pairs in LRL corpora. We propose a new Low resource, Adversarial, Cross-lingual (LAC) model for NMT. In terms of the adversary technique, LAC model consists of a generator and discriminator. The generator is a Seq2Seq model that produces the translations from source to target languages, while the discriminator measures the gap between machine and human translations. In addition, we introduce transfer learning on LAC model to help capture the features in rare resources because some languages share the same subject-verb-object grammatical structure. Rather than using the entire pretrained LAC model, we separately utilize the pretrained generator and discriminator. The pretrained discriminator exhibited better performance in all experiments. Experimental results demonstrate that the LAC model achieves higher Bilingual Evaluation Understudy (BLEU) scores and has good potential to augment LRL translations.
format	article
author	Mengtao Sun Hao Wang Mark Pasquine Ibrahim A. Hameed
author_facet	Mengtao Sun Hao Wang Mark Pasquine Ibrahim A. Hameed
author_sort	Mengtao Sun
title	Machine Translation in Low-Resource Languages by an Adversarial Neural Network
title_short	Machine Translation in Low-Resource Languages by an Adversarial Neural Network
title_full	Machine Translation in Low-Resource Languages by an Adversarial Neural Network
title_fullStr	Machine Translation in Low-Resource Languages by an Adversarial Neural Network
title_full_unstemmed	Machine Translation in Low-Resource Languages by an Adversarial Neural Network
title_sort	machine translation in low-resource languages by an adversarial neural network
publisher	MDPI AG
publishDate	2021
url	https://doaj.org/article/c042e7e12b1748c3830acce4f976d5e0
work_keys_str_mv	AT mengtaosun machinetranslationinlowresourcelanguagesbyanadversarialneuralnetwork AT haowang machinetranslationinlowresourcelanguagesbyanadversarialneuralnetwork AT markpasquine machinetranslationinlowresourcelanguagesbyanadversarialneuralnetwork AT ibrahimahameed machinetranslationinlowresourcelanguagesbyanadversarialneuralnetwork
_version_	1718413100582961152

Machine Translation in Low-Resource Languages by an Adversarial Neural Network

Ejemplares similares