English

TOD-BERT: Pre-trained Natural Language Understanding for Task-Oriented Dialogue

Computation and Language 2020-10-02 v3

Abstract

The underlying difference of linguistic patterns between general text and task-oriented dialogue makes existing pre-trained language models less useful in practice. In this work, we unify nine human-human and multi-turn task-oriented dialogue datasets for language modeling. To better model dialogue behavior during pre-training, we incorporate user and system tokens into the masked language modeling. We propose a contrastive objective function to simulate the response selection task. Our pre-trained task-oriented dialogue BERT (TOD-BERT) outperforms strong baselines like BERT on four downstream task-oriented dialogue applications, including intention recognition, dialogue state tracking, dialogue act prediction, and response selection. We also show that TOD-BERT has a stronger few-shot ability that can mitigate the data scarcity problem for task-oriented dialogue.

Keywords

Cite

@article{arxiv.2004.06871,
  title  = {TOD-BERT: Pre-trained Natural Language Understanding for Task-Oriented Dialogue},
  author = {Chien-Sheng Wu and Steven Hoi and Richard Socher and Caiming Xiong},
  journal= {arXiv preprint arXiv:2004.06871},
  year   = {2020}
}

Comments

EMNLP 2020 camera-ready

R2 v1 2026-06-23T14:51:42.580Z