Triangular Transfer: Freezing the Pivot for Triangular Machine Translation

被引:0
|
作者
Zhang, Meng [1 ]
Li, Liangyou [1 ]
Liu, Qun [1 ]
机构
[1] Huawei Noahs Ark Lab, 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
Triangular machine translation is a special case of low-resource machine translation where the language pair of interest has limited parallel data, but both languages have abundant parallel data with a pivot language. Naturally, the key to triangular machine translation is the successful exploitation of such auxiliary data. In this work, we propose a transfer-learning-based approach that utilizes all types of auxiliary data. As we train auxiliary source-pivot and pivot-target translation models, we initialize some parameters of the pivot side with a pre-trained language model and freeze them to encourage both translation models to work in the same pivot language space, so that they can be smoothly transferred to the source-target translation model. Experiments show that our approach can outperform previous ones.
引用
收藏
页码:644 / 650
页数:7
相关论文
共 50 条
  • [41] TRIANGULAR LATTICE FOLDINGS - A TRANSFER-MATRIX STUDY
    KANTOR, Y
    JARIC, MV
    EUROPHYSICS LETTERS, 1990, 11 (02): : 157 - 161
  • [42] Dual-domain deformation transfer for triangular meshes
    Zhao, Yong
    Pan, Bin
    Xiao, Chunxia
    Peng, Qunsheng
    COMPUTER ANIMATION AND VIRTUAL WORLDS, 2012, 23 (3-4) : 447 - 456
  • [43] Heat transfer to impinging round jets with triangular tabs
    Gao, N
    Sun, H
    Ewing, D
    INTERNATIONAL JOURNAL OF HEAT AND MASS TRANSFER, 2003, 46 (14) : 2557 - 2569
  • [44] Augmentation of heat transfer in a channel using a triangular prism
    Chattopadhyay, Himadri
    INTERNATIONAL JOURNAL OF THERMAL SCIENCES, 2007, 46 (05) : 501 - 505
  • [45] Improving Pivot Translation by Remembering the Pivot
    Miura, Akiva
    Neubig, Graham
    Sakti, Sakriani
    Toda, Tomoki
    Nakamura, Satoshi
    PROCEEDINGS OF THE 53RD ANNUAL MEETING OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS (ACL) AND THE 7TH INTERNATIONAL JOINT CONFERENCE ON NATURAL LANGUAGE PROCESSING (IJCNLP), VOL 2, 2015, : 573 - 577
  • [46] Power domination on triangular grids with triangular and hexagonal shape
    Prosenjit Bose
    Valentin Gledel
    Claire Pennarun
    Sander Verdonschot
    Journal of Combinatorial Optimization, 2020, 40 : 482 - 500
  • [47] Graded Triangular BasesGraded Triangular BasesJ. Brundan
    Jonathan Brundan
    Algebras and Representation Theory, 2025, 28 (1) : 239 - 280
  • [48] Triangular numbers whose sum of divisors is also triangular
    Iannucci, Douglas E.
    Luca, Florian
    ACTA ARITHMETICA, 2007, 129 (01) : 23 - 40
  • [49] Regularizing End-to-End Speech Translation with Triangular Decomposition Agreement
    Du, Yichao
    Zhang, Zhirui
    Wang, Weizhi
    Chen, Boxing
    Xie, Jun
    Xu, Tong
    THIRTY-SIXTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE / THIRTY-FOURTH CONFERENCE ON INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE / TWELVETH SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE, 2022, : 10590 - 10598
  • [50] Power domination on triangular grids with triangular and hexagonal shape
    Bose, Prosenjit
    Gledel, Valentin
    Pennarun, Claire
    Verdonschot, Sander
    JOURNAL OF COMBINATORIAL OPTIMIZATION, 2020, 40 (02) : 482 - 500