Neural machine translation: Challenges, progress and future

被引：25
作者：
Zhang, JiaJun [1 ,2 ]
Zong, ChengQing [1 ,2 ,3 ]
机构：
[1] CASIA, Natl Lab Pattern Recognit, Beijing 100190, Peoples R China
[2] Univ Chinese Acad Sci, Beijing 100190, Peoples R China
[3] CAS Ctr Excellence Brain Sci & Intelligence Techn, Shanghai 200031, Peoples R China
来源：
SCIENCE CHINA-TECHNOLOGICAL SCIENCES | 2020年 / 63卷 / 10期
基金：
中国国家自然科学基金;
关键词：
neural machine translation; Transformer; multimodal translation; low-resource translation; document translation; KNOWLEDGE; SEQUENCE; MODELS;
D O I：
10.1007/s11431-020-1632-x
中图分类号：
T [工业技术];
学科分类号：
08 ;
摘要：
Machine translation (MT) is a technique that leverages computers to translate human languages automatically. Nowadays, neural machine translation (NMT) which models direct mapping between source and target languages with deep neural networks has achieved a big breakthrough in translation performance and become the de facto paradigm of MT. This article makes a review of NMT framework, discusses the challenges in NMT, introduces some exciting recent progresses and finally looks forward to some potential future research trends.
引用
收藏
页码：2028 / 2050
页数：23
相关论文
共 177 条
[51]  
Emelianenko Dmitrii, 2019, ADV NEURAL INFORM PR, V32, P7698
[52]  
Faruqui M., 2014, Proc. 14th Conf. Eur. Chapter Assoc. Comput. Linguistics, Gothenburg, P462, DOI DOI 10.3115/V1/E14-1049
[53]  
Feng Y, 2019, P ANN M ASS COMP LIN, P1390
[54]   Multi-way, multilingual neural machine translation [J].
Firat, Orhan ;
Cho, Kyunghyun ;
Sankaran, Baskaran ;
Vural, Fatos T. Yarman ;
Bengio, Yoshua .
COMPUTER SPEECH AND LANGUAGE, 2017, 45 :236-252
[55]  
Gehring J, 2017, PR MACH LEARN RES, V70
[56]  
Gong Z., 2011, EMNLP, P909
[57]  
Graves A., 2006, P 23 INT C MACHINE L, P369
[58]  
Grissom II A, 2014, P 2014 C EMPIRICAL M, P1342, DOI [DOI 10.3115/V1/D14-1140, 10.3115/v1/D14-1140]
[59]  
Gu J, 2017, P 15 C EUR ASS COMP, P1053
[60]  
Gu J, 2018, P INT C LEARN REPR V
12345678910