Co-Regularized Adversarial Learning for Multi-Domain Text Classification

被引:0
|
作者
Wu, Yuan [1 ]
Inkpen, Diana [2 ]
El-Roby, Ahmed [1 ]
机构
[1] Carleton Univ, Ottawa, ON, Canada
[2] Univ Ottawa, Ottawa, ON, Canada
来源
INTERNATIONAL CONFERENCE ON ARTIFICIAL INTELLIGENCE AND STATISTICS, VOL 151 | 2022年 / 151卷
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
Multi-domain text classification (MDTC) aims to leverage all available resources from multiple domains to learn a predictive model that can generalize well on these domains. Recently, many MDTC methods adopt adversarial learning, shared-private paradigm, and entropy minimization to yield state-of-the-art results. However, these approaches face three issues: (1) Minimizing domain divergence can not fully guarantee the success of domain alignment; (2) Aligning marginal feature distributions can not fully guarantee the discriminability of the learned features; (3) Standard entropy minimization may make the predictions on unlabeled data over-confident, deteriorating the discriminability of the learned features. In order to address the above issues, we propose a co-regularized adversarial learning (CRAL) mechanism for MDTC. This approach constructs two diverse shared latent spaces, performs domain alignment in each of them, and punishes the disagreements of these two alignments with respect to the predictions on unlabeled data. Moreover, virtual adversarial training (VAT) with entropy minimization is incorporated to impose consistency regularization to the CRAL method. Experiments show that our model outperforms state-of-the-art methods on two MDTC benchmarks.
引用
收藏
页数:12
相关论文
共 50 条
  • [1] MIXUP REGULARIZED ADVERSARIAL NETWORKS FOR MULTI-DOMAIN TEXT CLASSIFICATION
    Wu, Yuan
    Inkpen, Diana
    El-Roby, Ahmed
    2021 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP 2021), 2021, : 7733 - 7737
  • [2] Dual Adversarial Co-Learning for Multi-Domain Text Classification
    Wu, Yuan
    Guo, Yuhong
    THIRTY-FOURTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE, THE THIRTY-SECOND INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE AND THE TENTH AAAI SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE, 2020, 34 : 6438 - 6445
  • [3] Flexible and Robust Co-Regularized Multi-Domain Graph Clustering
    Cheng, Wei
    Zhang, Xiang
    Guo, Zhishan
    Wu, Yubao
    Sullivan, Patrick F.
    Wang, Wei
    19TH ACM SIGKDD INTERNATIONAL CONFERENCE ON KNOWLEDGE DISCOVERY AND DATA MINING (KDD'13), 2013, : 320 - 328
  • [4] Learning Multi-Domain Adversarial Neural Networks for Text Classification
    Ding, Xiao
    Shi, Qiankun
    Cai, Bibo
    Liu, Ting
    Zhao, Yanyan
    Ye, Qiang
    IEEE ACCESS, 2019, 7 : 40323 - 40332
  • [5] DaCon: Multi-Domain Text Classification Using Domain Adversarial Contrastive Learning
    Dai, Yingjun
    El-Roby, Ahmed
    ARTIFICIAL NEURAL NETWORKS AND MACHINE LEARNING, ICANN 2023, PT V, 2023, 14258 : 40 - 52
  • [6] CGC: A Flexible and Robust Approach to Integrating Co-Regularized Multi-Domain Graph for Clustering
    Cheng, Wei
    Guo, Zhishan
    Zhang, Xiang
    Wang, Wei
    ACM TRANSACTIONS ON KNOWLEDGE DISCOVERY FROM DATA, 2016, 10 (04)
  • [7] Co-Regularized PLSA for Multi-Modal Learning
    Wang, Xin
    Chang, Ming-Ching
    Ying, Yiming
    Lyu, Siwei
    THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE, 2016, : 2166 - 2172
  • [8] Margin Discrepancy-Based Adversarial Training for Multi-Domain Text Classification
    Wu, Yuan
    NATURAL LANGUAGE PROCESSING AND CHINESE COMPUTING, PT III, NLPCC 2024, 2025, 15361 : 170 - 182
  • [9] Co-regularized Alignment for Unsupervised Domain Adaptation
    Kumar, Abhishek
    Sattigeri, Prasanna
    Wadhawan, Kahini
    Karlinsky, Leonid
    Feris, Rogerio
    Freeman, William T.
    Wornell, Gregory
    ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 31 (NIPS 2018), 2018, 31
  • [10] Co-Regularized Least Square Regression for Multi-View Multi-Class Classification
    Lan, Chao
    Deng, Yujie
    Li, Xiaoli
    Huan, Jun
    2016 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS (IJCNN), 2016, : 342 - 347