Dual Learning for Semi-Supervised Natural Language Understanding

被引:22
|
作者
Zhu, Su [1 ,2 ]
Cao, Ruisheng [1 ,2 ]
Yu, Kai [1 ,2 ]
机构
[1] Shanghai Jiao Tong Univ, Dept Comp Sci & Engn, AI Inst, Shanghai 200240, Peoples R China
[2] Shanghai Jiao Tong Univ, MoE Key Lab Artificial Intelligence, AI Inst, Shanghai 200240, Peoples R China
关键词
Task analysis; Semantics; Natural languages; Supervised learning; Neural networks; Annotations; Speech processing; Natural language understanding (NLU); semi-supervised learning; dual learning; slot filling; intent detection; RECURRENT NEURAL-NETWORKS; MODELS;
D O I
10.1109/TASLP.2020.3001684
中图分类号
O42 [声学];
学科分类号
070206 ; 082403 ;
摘要
Natural language understanding (NLU) converts sentences into structured semantic forms. The paucity of annotated training samples is still a fundamental challenge of NLU. To solve this data sparsity problem, previous work based on semi-supervised learning mainly focuses on exploiting unlabeled sentences. In this work, we introduce a dual task of NLU, semantic-to-sentence generation (SSG), and propose a new framework for semi-supervised NLU with the corresponding dual model. The framework is composed of dual pseudo-labeling and dual learning method, which enables an NLU model to make full use of data (labeled and unlabeled) through a closed-loop of the primal and dual tasks. By incorporating the dual task, the framework can exploit pure semantic forms as well as unlabeled sentences, and further improve the NLU and SSG models iteratively in the closed-loop. The proposed approaches are evaluated on two public datasets (ATIS and SNIPS). Experiments in the semi-supervised setting show that our methods can outperform various baselines significantly, and extensive ablation studies are conducted to verify the effectiveness of our framework. Finally, our method can also achieve the state-of-the-art performance on the two datasets in the supervised setting.
引用
收藏
页码:1936 / 1947
页数:12
相关论文
共 50 条
  • [31] A survey on semi-supervised learning
    Jesper E. van Engelen
    Holger H. Hoos
    [J]. Machine Learning, 2020, 109 : 373 - 440
  • [32] Semi-supervised Sequence Learning
    Dai, Andrew M.
    Le, Quoc V.
    [J]. ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 28 (NIPS 2015), 2015, 28
  • [33] Semi-supervised learning by disagreement
    Zhi-Hua Zhou
    Ming Li
    [J]. Knowledge and Information Systems, 2010, 24 : 415 - 439
  • [34] Semi-Supervised Incremental Learning
    Bouchachia, Abdelhamid
    Prossegger, Markus
    Duman, Hakan
    [J]. 2010 IEEE INTERNATIONAL CONFERENCE ON FUZZY SYSTEMS (FUZZ-IEEE 2010), 2010,
  • [35] Deep Semi-Supervised Learning
    Hailat, Zeyad
    Komarichev, Artem
    Chen, Xue-Wen
    [J]. 2018 24TH INTERNATIONAL CONFERENCE ON PATTERN RECOGNITION (ICPR), 2018, : 2154 - 2159
  • [36] Semi-Supervised Learning by Disagreement
    Zhou, Zhi-Hua
    [J]. 2008 IEEE INTERNATIONAL CONFERENCE ON GRANULAR COMPUTING, VOLS 1 AND 2, 2008, : 93 - 93
  • [37] Reliable Semi-supervised Learning
    Shao, Junming
    Huang, Chen
    Yang, Qinli
    Luo, Guangchun
    [J]. 2016 IEEE 16TH INTERNATIONAL CONFERENCE ON DATA MINING (ICDM), 2016, : 1197 - 1202
  • [38] Semi-supervised Learning with Transfer Learning
    Zhou, Huiwei
    Zhang, Yan
    Huang, Degen
    Li, Lishuang
    [J]. CHINESE COMPUTATIONAL LINGUISTICS AND NATURAL LANGUAGE PROCESSING BASED ON NATURALLY ANNOTATED BIG DATA, 2013, 8208 : 109 - 119
  • [39] Semi-supervised learning with dropouts
    Abhishek
    Yadav, Rakesh Kumar
    Verma, Shekhar
    [J]. EXPERT SYSTEMS WITH APPLICATIONS, 2023, 215
  • [40] PRIVILEGED SEMI-SUPERVISED LEARNING
    Chen, Xingyu
    Gong, Chen
    Ma, Chao
    Huang, Xiaolin
    Yang, Jie
    [J]. 2018 25TH IEEE INTERNATIONAL CONFERENCE ON IMAGE PROCESSING (ICIP), 2018, : 2999 - 3003