Continual Graph Convolutional Networks for Text Classification

Nov. 19^th, 2022: Got accepted by AAAI ‘23!

Tiandeng WU^*, Qijiong LIU^*, Yao HUANG, Yi CAO, Xiao-Ming WU^#, and Jiandong DING^#
*Equal contribution (co-first authors). Author ordering determined by dice rolling.
[Code] [Paper]

Abstract

To capture global non-consecutive and long-distance semantic information, graph convolutional network (GCN) has been widely used for text classification. While GCN-based methods have achieved great success in offline evaluations, they usually construct fixed document-token graphs and cannot perform inference on new documents. It is still a challenge to apply GCNs in online systems which need to infer continual text data. In this work, we present a Continual GCN model, short as ContGCN, to generalize inferences from observed documents to unobserved documents. Concretely, we propose a novel global-token-local-document paradigm to dynamically update the document-token graph in every batch for any online system during both training and testing phases. Moreover, we design an occurrence memory module and a self-supervised contrastive learning objective to update the proposed ContGCN in any online system in a label-free manner. Extensive offline experiments conducted on five public datasets demonstrate that our proposed ContGCN can significantly improve inference quality. A 3-month A/B test on our internal online system shows ContGCN achieves 8.86% performance gain compared with state-of-the-art methods.

Citation

@inproceedings{wu2023contgcn,
    title = "Continual Graph Convolutional Networks for Text Classification",
    author = "Wu, Tiandeng  and
      Liu, Qijiong  and
      Cao, Yi  and
      Huang, Yao  and
      Wu, Xiaoming and
      Ding, Jiandong",
    booktitle = "Proceedings of the 37th AAAI Conference on Artificial Intelligence",
    month = feb,
    year = "2023",
    address = "Washington, DC, United States"
}