The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. ex. Some numerals are expressed as "XNUMX".
Copyrights notice
The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. Copyrights notice
Como abordagem dominante no campo da tradução automática, a tradução automática neural (NMT) alcançou grandes melhorias em muitas línguas de fontes ricas, mas o desempenho da NMT para línguas de poucos recursos ainda não é muito bom. Este artigo utiliza tecnologia de aprimoramento de dados para construir corpus pseudo-paralelo mongol-chinês, a fim de melhorar a capacidade de tradução do modelo de tradução mongol-chinês. Experimentos mostram que os métodos acima podem melhorar a capacidade de tradução do modelo de tradução. Finalmente, obtém-se um modelo de tradução treinado com corpus pseudo-paralelo em larga escala e integrado com tecnologia de aprimoramento de dados de contexto suave, e seu valor BLEU é 39.3.
Qing-dao-er-ji REN
Inner Mongolia University of Technology
Yuan LI
Inner Mongolia University of Technology
Shi BAO
Inner Mongolia University of Technology
Yong-chao LIU
Inner Mongolia University of Technology
Xiu-hong CHEN
Hohhot, Inner Mongolia
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copiar
Qing-dao-er-ji REN, Yuan LI, Shi BAO, Yong-chao LIU, Xiu-hong CHEN, "Research on Mongolian-Chinese Translation Model Based on Transformer with Soft Context Data Augmentation Technique" in IEICE TRANSACTIONS on Fundamentals,
vol. E105-A, no. 5, pp. 871-876, May 2022, doi: 10.1587/transfun.2021EAP1121.
Abstract: As the mainstream approach in the field of machine translation, neural machine translation (NMT) has achieved great improvements on many rich-source languages, but performance of NMT for low-resource languages ae not very good yet. This paper uses data enhancement technology to construct Mongolian-Chinese pseudo parallel corpus, so as to improve the translation ability of Mongolian-Chinese translation model. Experiments show that the above methods can improve the translation ability of the translation model. Finally, a translation model trained with large-scale pseudo parallel corpus and integrated with soft context data enhancement technology is obtained, and its BLEU value is 39.3.
URL: https://global.ieice.org/en_transactions/fundamentals/10.1587/transfun.2021EAP1121/_p
Copiar
@ARTICLE{e105-a_5_871,
author={Qing-dao-er-ji REN, Yuan LI, Shi BAO, Yong-chao LIU, Xiu-hong CHEN, },
journal={IEICE TRANSACTIONS on Fundamentals},
title={Research on Mongolian-Chinese Translation Model Based on Transformer with Soft Context Data Augmentation Technique},
year={2022},
volume={E105-A},
number={5},
pages={871-876},
abstract={As the mainstream approach in the field of machine translation, neural machine translation (NMT) has achieved great improvements on many rich-source languages, but performance of NMT for low-resource languages ae not very good yet. This paper uses data enhancement technology to construct Mongolian-Chinese pseudo parallel corpus, so as to improve the translation ability of Mongolian-Chinese translation model. Experiments show that the above methods can improve the translation ability of the translation model. Finally, a translation model trained with large-scale pseudo parallel corpus and integrated with soft context data enhancement technology is obtained, and its BLEU value is 39.3.},
keywords={},
doi={10.1587/transfun.2021EAP1121},
ISSN={1745-1337},
month={May},}
Copiar
TY - JOUR
TI - Research on Mongolian-Chinese Translation Model Based on Transformer with Soft Context Data Augmentation Technique
T2 - IEICE TRANSACTIONS on Fundamentals
SP - 871
EP - 876
AU - Qing-dao-er-ji REN
AU - Yuan LI
AU - Shi BAO
AU - Yong-chao LIU
AU - Xiu-hong CHEN
PY - 2022
DO - 10.1587/transfun.2021EAP1121
JO - IEICE TRANSACTIONS on Fundamentals
SN - 1745-1337
VL - E105-A
IS - 5
JA - IEICE TRANSACTIONS on Fundamentals
Y1 - May 2022
AB - As the mainstream approach in the field of machine translation, neural machine translation (NMT) has achieved great improvements on many rich-source languages, but performance of NMT for low-resource languages ae not very good yet. This paper uses data enhancement technology to construct Mongolian-Chinese pseudo parallel corpus, so as to improve the translation ability of Mongolian-Chinese translation model. Experiments show that the above methods can improve the translation ability of the translation model. Finally, a translation model trained with large-scale pseudo parallel corpus and integrated with soft context data enhancement technology is obtained, and its BLEU value is 39.3.
ER -