Non-Autoregressive Machine Translation with Auxiliary Regularization

被引:0
|
作者
Wang, Yiren [1 ]
Tian, Fei [2 ]
He, Di [3 ]
Qin, Tao [2 ]
Zhai, ChengXiang [1 ]
Liu, Tie-Yan [2 ]
机构
[1] Univ Illinois, Urbana, IL 61801 USA
[2] Microsoft Res, Beijing, Peoples R China
[3] Peking Univ, Key Lab Machine Percept, MOE, Sch EECS, Beijing, Peoples R China
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
As a new neural machine translation approach, Non-Autoregressive machine Translation (NAT) has attracted attention recently due to its high efficiency in inference. However, the high efficiency has come at the cost of not capturing the sequential dependency on the target side of translation, which causes NAT to suffer from two kinds of translation errors: 1) repeated translations (due to indistinguishable adjacent decoder hidden states), and 2) incomplete translations (due to incomplete transfer of source side information via the decoder hidden states). In this paper, we propose to address these two problems by improving the quality of decoder hidden representations via two auxiliary regularization terms in the training process of an NAT model. First, to make the hidden states more distinguishable, we regularize the similarity between consecutive hidden states based on the corresponding target tokens. Second, to force the hidden states to contain all the information in the source sentence, we leverage the dual nature of translation tasks (e.g., English to German and German to English) and minimize a backward reconstruction error to ensure that the hidden states of the NAT decoder are able to recover the source side sentence. Extensive experiments conducted on several benchmark datasets show that both regularization strategies are effective and can alleviate the issues of repeated translations and incomplete translations in NAT models. The accuracy of NAT models is therefore improved significantly over the state-of-the-art NAT models with even better efficiency for inference.
引用
收藏
页码:5377 / 5384
页数:8
相关论文
共 50 条
  • [41] Multi-Task Learning with Shared Encoder for Non-Autoregressive Machine Translation
    Hao, Yongchang
    He, Shilin
    Jiao, Wenxiang
    Tu, Zhaopeng
    Lyu, Michael R.
    Wang, Xing
    [J]. 2021 CONFERENCE OF THE NORTH AMERICAN CHAPTER OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS: HUMAN LANGUAGE TECHNOLOGIES (NAACL-HLT 2021), 2021, : 3989 - 3996
  • [42] Non-Monotonic Latent Alignments for CTC-Based Non-Autoregressive Machine Translation
    Shao, Chenze
    Feng, Yang
    [J]. ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 35 (NEURIPS 2022), 2022,
  • [43] End-to-End Non-Autoregressive Neural Machine Translation with Connectionist Temporal Classification
    Libovicky, Jindrich
    Helcl, Jindrich
    [J]. 2018 CONFERENCE ON EMPIRICAL METHODS IN NATURAL LANGUAGE PROCESSING (EMNLP 2018), 2018, : 3016 - 3021
  • [44] A Non-Autoregressive Neural Machine Translation Model With Iterative Length Update of Target Sentence
    Lim, Yeon-Soo
    Park, Eun-Ju
    Song, Hyun-Je
    Park, Seong-Bae
    [J]. IEEE ACCESS, 2022, 10 : 43341 - 43350
  • [45] AligNART: Non-autoregressive Neural Machine Translation by Jointly Learning to Estimate Alignment and Translate
    Song, Jongyoon
    Kim, Sungwon
    Yoon, Sungroh
    [J]. 2021 CONFERENCE ON EMPIRICAL METHODS IN NATURAL LANGUAGE PROCESSING (EMNLP 2021), 2021, : 1 - 14
  • [46] Improving Non-Autoregressive Machine Translation Using Sentence-Level Semantic Agreement
    Wang, Shuheng
    Huang, Heyan
    Shi, Shumin
    [J]. APPLIED SCIENCES-BASEL, 2022, 12 (10):
  • [47] Non-Autoregressive Translation by Learning Target Categorical Codes
    Bao, Yu
    Huang, Shujian
    Xiao, Tong
    Wang, Dongqi
    Dai, Xinyu
    Chen, Jiajun
    [J]. 2021 CONFERENCE OF THE NORTH AMERICAN CHAPTER OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS: HUMAN LANGUAGE TECHNOLOGIES (NAACL-HLT 2021), 2021, : 5749 - 5759
  • [48] CTC-based Non-autoregressive Speech Translation
    Xu, Chen
    Liu, Xiaoqian
    Liu, Xiaowen
    Sun, Qingxuan
    Zhang, Yuhao
    Yang, Murun
    Dong, Qianqian
    Ko, Tom
    Wang, Mingxuan
    Xiao, Tong
    Ma, Anxiang
    Zhu, Jingbo
    [J]. PROCEEDINGS OF THE 61ST ANNUAL MEETING OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS (ACL 2023): LONG PAPERS, VOL 1, 2023, : 13321 - 13339
  • [49] Jointly Masked Sequence-to-Sequence Model for Non-Autoregressive Neural Machine Translation
    Guo, Junliang
    Xu, Linli
    Chen, Enhong
    [J]. 58TH ANNUAL MEETING OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS (ACL 2020), 2020, : 376 - 385
  • [50] Learning to Recover from Multi-Modality Errors for Non-Autoregressive Neural Machine Translation
    Ran, Qiu
    Lin, Yankai
    Li, Peng
    Zhou, Jie
    [J]. 58TH ANNUAL MEETING OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS (ACL 2020), 2020, : 3059 - 3069