Self-training Improves Pre-training for Few-shot Learning in Task-oriented Dialog Systems

被引:0
|
作者
Mi, Fei [1 ]
Zhou, Wanhao [2 ]
Cai, Fengyu [2 ]
Kong, Lingjing [2 ]
Huang, Minlie [3 ]
Faltings, Boi [2 ]
机构
[1] Huawei Noahs Ark Lab, Hong Kong, Peoples R China
[2] Ecole Polytech Fed Lausanne, LIA, Lausanne, Switzerland
[3] Tsinghua Univ, CoAI, DCST, 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 the labeling cost for different modules in task-oriented dialog (ToD) systems is expensive, a major challenge is to train different modules with the least amount of labeled data. Recently, large-scale pre-trained language models, have shown promising results for few-shot learning in ToD. In this paper, we devise a self-training approach to utilize the abundant unlabeled dialog data to further improve state-of-the-art pre-trained models in few-shot learning scenarios for ToD systems. Specifically, we propose a self-training approach that iteratively labels the most confident unlabeled data to train a stronger Student model. Moreover, a new text augmentation technique (GradAug) is proposed to better train the Student by replacing non-crucial tokens using a masked language model. We conduct extensive experiments and present analyses on four downstream tasks in ToD, including intent classification, dialog state tracking, dialog act prediction, and response selection. Empirical results demonstrate that the proposed self-training approach consistently improves state-of-the-art pre-trained models (BERT, ToD-BERT) when only a small number of labeled data are available.
引用
收藏
页码:1887 / 1898
页数:12
相关论文
共 50 条
  • [1] CINS: Comprehensive Instruction for Few-Shot Learning in Task-Oriented Dialog Systems
    Mi, Fei
    Wang, Yasheng
    Li, Yitong
    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, : 11076 - 11084
  • [2] STraTA: Self-Training with Task Augmentation for Better Few-shot Learning
    Tu Vu
    Minh-Thang Luong
    Le, Quoc, V
    Simon, Grady
    Iyyer, Mohit
    2021 CONFERENCE ON EMPIRICAL METHODS IN NATURAL LANGUAGE PROCESSING (EMNLP 2021), 2021, : 5715 - 5731
  • [3] Unified Dialog Model Pre-training for Task-Oriented Dialog Understanding and Generation
    He, Wanwei
    Dai, Yinpei
    Yang, Min
    Sun, Jian
    Huang, Fei
    Si, Luo
    Li, Yongbin
    PROCEEDINGS OF THE 45TH INTERNATIONAL ACM SIGIR CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION RETRIEVAL (SIGIR '22), 2022, : 187 - 200
  • [4] Self-training with Few-shot Rationalization
    Bhat, Meghana Moorthy
    Sordoni, Alessandro
    Mukherjee, Subhabrata
    2021 CONFERENCE ON EMPIRICAL METHODS IN NATURAL LANGUAGE PROCESSING (EMNLP 2021), 2021, : 10702 - 10712
  • [5] Self-training improves few-shot learning in legal artificial intelligence tasks
    Zhou, Yulin
    Qin, Yongbin
    Huang, Ruizhang
    Chen, Yanping
    Lin, Chuan
    Zhou, Yuan
    ARTIFICIAL INTELLIGENCE AND LAW, 2024,
  • [6] PRAL: A Tailored Pre-Training Model for Task-Oriented Dialog Generation
    Gu, Jing
    Wu, Qingyang
    Wu, Chongruo
    Shi, Weiyan
    Yu, Zhou
    ACL-IJCNLP 2021: THE 59TH ANNUAL MEETING OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS AND THE 11TH INTERNATIONAL JOINT CONFERENCE ON NATURAL LANGUAGE PROCESSING, VOL 2, 2021, : 305 - 313
  • [7] Few-shot Natural Language Generation for Task-Oriented Dialog
    Peng, Baolin
    Zhu, Chenguang
    Li, Chunyuan
    Li, Xiujun
    Li, Jinchao
    Zeng, Michael
    Gao, Jianfeng
    FINDINGS OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS, EMNLP 2020, 2020, : 172 - 182
  • [8] Revisiting Self-Training for Few-Shot Learning of Language Model
    Chen, Yiming
    Zhang, Yan
    Zhang, Chen
    Lee, Grandee
    Cheng, Ran
    Li, Haizhou
    2021 CONFERENCE ON EMPIRICAL METHODS IN NATURAL LANGUAGE PROCESSING (EMNLP 2021), 2021, : 9125 - 9135
  • [9] Omni-Training: Bridging Pre-Training and Meta-Training for Few-Shot Learning
    Shu, Yang
    Cao, Zhangjie
    Gao, Jinghan
    Wang, Jianmin
    Yu, Philip S.
    Long, Mingsheng
    IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2023, 45 (12) : 15275 - 15291
  • [10] Multitask Pre-training of Modular Prompt for Chinese Few-Shot Learning
    Sun, Tianxiang
    He, Zhengfu
    Zhu, Qin
    Qiu, Xipeng
    Huang, Xuanjing
    PROCEEDINGS OF THE 61ST ANNUAL MEETING OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS (ACL 2023): LONG PAPERS, VOL 1, 2023, : 11156 - 11172