Multilingual Machine Translation Systems at WAT 2021: One-to-Many and Many-to-One Transformer based NMT

被引：0

作者：

Mhaskar, Shivam ^{[1
]}

Jain, Aditya ^{[1
]}

Banerjee, Aakash ^{[1
]}

Bhattacharyya, Pushpak ^{[1
]}

机构：

[1] Indian Inst Technol, Dept Comp Sci & Engn, Mumbai, Maharashtra, India

来源：

WAT 2021: THE 8TH WORKSHOP ON ASIAN TRANSLATION | 2021年

关键词：

D O I：

暂无

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

In this paper, we present the details of the systems that we have submitted for the WAT 2021 MultiIndicMT: An Indic Language Multilingual Task. We have submitted two separate multilingual NMT models: one for English to 10 Indic languages and another for 10 Indic languages to English. We discuss the implementation details of two separate multilingual NMT approaches, namely one-to-many and many-to-one, that makes use of a shared decoder and a shared encoder, respectively. From our experiments, we observe that the multilingual NMT systems outperforms the bilingual baseline MT systems for each of the language pairs under consideration.

引用

页码：233 / 237

页数：5

共 19 条

[1] [Anonymous], 2014, Advances in neural information processing systems
[2] Bahdanau D, 2016, Arxiv, DOI arXiv:1409.0473
[3] Adequacy-Fluency Metrics: Evaluating MT in the Continuous Space Model Framework
Banchs, Rafael E.
D'Haro, Luis F.
Li, Haizhou
[J]. IEEE-ACM TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE PROCESSING, 2015, 23 (03) : 472 - 482
[4] Cho K., 2014, P SSST 8 8 WORKSH SY, P103, DOI [10.3115/v1/w14-4012, DOI 10.3115/V1/W14-4012, DOI 10.3115/V1/D14-1179]
[5] Dabre Raj, 2019, ABS190505395 CORR
[6] Dong DX, 2015, PROCEEDINGS OF THE 53RD ANNUAL MEETING OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS AND THE 7TH INTERNATIONAL JOINT CONFERENCE ON NATURAL LANGUAGE PROCESSING, VOL 1, P1723
[7] Firat O, 2016, P 2016 C N AM CHAPT, P866, DOI [DOI 10.1016/J.CSL.2016.10.006, 10.18653/v1/n16-1101]
[8] Ha Thanh-Le, 2016, P 13 INT WORKSH SPOK
[9] Isozaki H., 2010, P 2010 C EMP METH NA, P944
[10] Johnson M., 2017, Transactions of the Association for Computational Linguistics, V5, P339, DOI 10.1162/tacla00065

← 1 2 →