Learning Multi-Task Communication with Message Passing for Sequence Learning

Pengfei Liu, Jie Fu, Yue Dong, Xipeng Qiu, Jackie Chi Kit Cheung

AAAI 2019 pp. 4360-4367

doi:10.1609/AAAI.V33I01.33014360 /aaai/2019/liu2019aaai-learning/

Abstract

We present two architectures for multi-task learning with neural sequence models. Our approach allows the relationships between different tasks to be learned dynamically, rather than using an ad-hoc pre-defined structure as in previous work. We adopt the idea from message-passing graph neural networks, and propose a general graph multi-task learning framework in which different tasks can communicate with each other in an effective and interpretable way. We conduct extensive experiments in text classification and sequence labelling to evaluate our approach on multi-task learning and transfer learning. The empirical results show that our models not only outperform competitive baselines, but also learn interpretable and transferable patterns across tasks.

PDF AAAI Semantic Scholar

Cite

Text

Liu et al. "Learning Multi-Task Communication with Message Passing for Sequence Learning." AAAI Conference on Artificial Intelligence, 2019. doi:10.1609/AAAI.V33I01.33014360

Markdown

[Liu et al. "Learning Multi-Task Communication with Message Passing for Sequence Learning." AAAI Conference on Artificial Intelligence, 2019.](https://mlanthology.org/aaai/2019/liu2019aaai-learning/) doi:10.1609/AAAI.V33I01.33014360

BibTeX

@inproceedings{liu2019aaai-learning,
  title     = {{Learning Multi-Task Communication with Message Passing for Sequence Learning}},
  author    = {Liu, Pengfei and Fu, Jie and Dong, Yue and Qiu, Xipeng and Cheung, Jackie Chi Kit},
  booktitle = {AAAI Conference on Artificial Intelligence},
  year      = {2019},
  pages     = {4360-4367},
  doi       = {10.1609/AAAI.V33I01.33014360},
  url       = {https://mlanthology.org/aaai/2019/liu2019aaai-learning/}
}