Elamur Research⌕
☰
home / papers / 2020 / Multi-Domain Dialogue Acts and Response Co-Generation

Multi-Domain Dialogue Acts and Response Co-Generation

Summary

Task-oriented dialogue systems usually produce a reply in two steps. They first predict dialogue acts (domain-action-slot triples) and then generate text conditioned on those acts. The authors identify two weaknesses in this pipeline. Classifying each act separately ignores the hierarchical structure that acts share across domains. Compressing all acts into one static vector also stops the text generator from focusing on the act relevant to each part of the reply.

They propose MarCo, a transformer-based model with a shared encoder over dialogue history and database attributes. The encoder feeds two decoders. One emits the acts as a token sequence ordered by domain, action and slot. The other writes the response while attending dynamically to the act decoder's hidden states. Both decoders are trained jointly, and their task weights are learned automatically through an uncertainty loss instead of being tuned by hand.

On MultiWOZ, MarCo outperforms several state-of-the-art systems, including HDSA, in both automatic and human evaluation. The work shows that treating act prediction as generation, tightly coupled with response decoding, yields more relevant and informative replies.

Key points

  • Generating acts as a sequence lets different domains share common slots (e.g., area). This shrinks the act search space compared with one-hot or binary-classifier schemes such as HDSA.
  • The response generator uses dynamic act attention to focus on different acts for different sub-spans of the reply. Joint training also reduces the error propagation typical of pipelines.
  • The uncertainty loss (Kendall et al., 2018) was compared on the development set against a weighted-sum loss whose weight was swept from 0 to 1 in steps of 0.1. The uncertainty loss was consistently better. Exact automatic scores fall in the part of the paper missing from the extracted text.
  • The human study used 100 test dialogues (742 turns), each rated by 3 judges for completion and readability:
    • MarCo beat both HDSA and the human references on completion.
    • It tied with HDSA in 94% of readability judgments.
    • It trailed the human references on readability, mainly because of token repetition.
  • Code and human-evaluation annotations are available at github.com/InitialBug/MarCo-Dialog.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

Fields: Linguistics & languages

Topics: Speech act theory

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2020.acl-main.638 · arXiv 2004.12363

@inproceedings{wang2020multi,
  title = {{Multi-Domain Dialogue Acts and Response Co-Generation}},
  author = {Wang, Kai and Tian, Junfeng and Wang, Rui and Quan, Xiaojun and Yu, Jianxing},
  booktitle = {Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics},
  year = {2020},
  doi = {10.18653/v1/2020.acl-main.638},
  eprint = {2004.12363},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.18653/v1/2020.acl-main.638}
}

References

32 references, 6 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Antoine Bordes, Y-Lan Boureau, and Jason Weston. 2017. Learning end-to-end goal-oriented dialog. In 5th International Conference on Learning Representations, ICLR 2017, Toulon, France, April 24-26, 2017, Conference Track Proceedings. find
  2. Paweł Budzianowski, Tsung-Hsien Wen, Bo-Hsiang Tseng, Iñigo Casanueva, Ultes Stefan, Ramadan Osman, and Milica Gašić. 2018. Multiwoz - a large-scale multi-domain wizard-of-oz dataset for task-oriented dialogue modelling. In Proceedings of the 2018 Conference on Empirical Methods in Natural Langua… find
  3. Wenhu Chen, Jianshu Chen, Pengda Qin, Xifeng Yan, and William Yang Wang. 2019. Semantically conditioned dialog response generation via hierarchical disentangled self-attention. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pages 3696–3709, Florence, Ital… find
  4. Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. Bert: Pre-training of deep bidirectional transformers for language understanding. In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies,… find
  5. Mihail Eric, Rahul Goel, Shachi Paul, Abhishek Sethi, Sanchit Agarwal, Shuyag Gao, and Dilek Hakkani-Tur. 2019. Multiwoz 2.1: Multi-domain dialogue state corrections and state tracking baselines. arXiv preprint arXiv:1907.01669. find
  6. Sonal Gupta, Rushin Shah, Mrinal Mohit, Anuj Kumar, and Mike Lewis. 2018. Semantic parsing for task oriented dialog using hierarchical representations. In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, pages 2787–2792, Brussels, Belgium. Association for Compu… find
  7. Zhuoxuan Jiang, Ziming Huang, Dong Sheng Li, and Xianling Mao. 2019. Dialogact2vec: Towards end-to-end dialogue agent by multi-task representation learning. arXiv preprint arXiv:1911.04088. — [arXiv:1911.04088](https://arxiv.org/abs/1911.04088) find
  8. Alex Kendall, Yarin Gal, and Roberto Cipolla. 2018. Multi-task learning using uncertainty to weigh losses for scene geometry and semantics. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 7482–7491. find
  9. Diederik P. Kingma and Jimmy Ba. 2015. Adam: A method for stochastic optimization. In 3rd International Conference on Learning Representations, ICLR 2015, San Diego, CA, USA, May 7-9, 2015, Conference Track Proceedings. find
  10. Jiwei Li, Will Monroe, Alan Ritter, Dan Jurafsky, Michel Galley, and Jianfeng Gao. 2016a. Deep reinforcement learning for dialogue generation. In Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing, EMNLP 2016, pages 1192–1202. find our card
  11. Jiwei Li, Will Monroe, Alan Ritter, Dan Jurafsky, Michel Galley, and Jianfeng Gao. 2016b. Deep reinforcement learning for dialogue generation. In Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing, pages 1192– 1202. find our card
  12. Bing Liu and Ian Lane. 2017. An end-to-end trainable neural network model with belief tracking for task-oriented dialog. In Interspeech 2017, 18th Annual Conference of the International Speech Communication Association, Stockholm, Sweden, August 20-24, 2017, pages 2506–2510. ISCA. find
  13. Shikib Mehri, Tejas Srinivasan, and Maxine Eskenazi. 2019. Structured fusion networks for dialog. arXiv preprint arXiv:1907.10016. find
  14. Haitao Mi, Baskaran Sankaran, Zhiguo Wang, and Abe Ittycheriah. 2016. Coverage embedding models for neural machine translation. arXiv preprint arXiv:1605.03148. find
  15. Nikola Mrkšić, Diarmuid Ó Séaghdha, Tsung-Hsien Wen, Blaise Thomson, and Steve Young. 2017. Neural belief tracker: Data-driven dialogue state tracking. In Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 1777–1788, Vancouver,… find
  16. Kishore Papineni, Salim Roukos, Todd Ward, and Wei-Jing Zhu. 2002. Bleu: a method for automatic evaluation of machine translation. In Proceedings of the 40th annual meeting on association for computational linguistics, pages 311–318. Association for Computational Linguistics. find our card
  17. Romain Paulus, Caiming Xiong, and Richard Socher. 2018. A deep reinforced model for abstractive summarization. In International Conference on Learning Representations. find our card
  18. Jiahuan Pei, Pengjie Ren, and Maarten de Rijke. 2019. A modular task-oriented dialogue system using a neural mixture-of-experts. arXiv preprint arXiv:1907.05346. — [arXiv:1907.05346](https://arxiv.org/abs/1907.05346) find
  19. Abhinav Rastogi, Raghav Gupta, and Dilek Hakkani-Tur. 2018. Multi-task learning for joint language understanding and dialogue state tracking. In Proceedings of the 19th Annual SIGdial Meeting on Discourse and Dialogue, pages 376–384. find
  20. Iulian V Serban, Chinnadhurai Sankar, Mathieu Germain, Saizheng Zhang, Zhouhan Lin, Sandeep Subramanian, Taesup Kim, Michael Pieper, Sarath Chandar, Nan Rosemary Ke, et al. 2017. A deep reinforcement learning chatbot. arXiv preprint arXiv:1709.02349. — [arXiv:1709.02349](https://arxiv.org/abs/1709.02349) find
  21. Anuroop Sriram, Heewoo Jun, Sanjeev Satheesh, and Adam Coates. 2018. Cold fusion: Training seq2seq models together with language models. In 6th International Conference on Learning Representations, ICLR 2018, Vancouver, BC, Canada, April 30 - May 3, 2018, Workshop Track Proceedings. find
  22. Pei-Hao Su, Paweł Budzianowski, Stefan Ultes, Milica Gasic, and Steve Young. 2017. Sample-efficient actor-critic reinforcement learning with supervised data for dialogue management. In Proceedings of the 18th Annual SIGdial Meeting on Discourse and Dialogue, pages 147–157. find
  23. Shang-Yu Su, Kai-Ling Lo, Yi-Ting Yeh, and Yun-Nung Chen. 2018. Natural language generation by hierarchical decoding with linguistic patterns. In Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 2 (… find
  24. Yizhe Zhang, Siqi Sun, Michel Galley, Yen-Chun Chen, Chris Brockett, Xiang Gao, Jianfeng Gao, Jingjing Liu, and Bill Dolan. 2019. Dialogpt: Large-scale generative pre-training for conversational response generation. arXiv preprint arXiv:1911.00536. find our card
  25. Tiancheng Zhao, Kaige Xie, and Maxine Eskenazi. 2019. Rethinking action spaces for reinforcement learning in end-to-end dialog agents with latent variable models. In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Tech… find
  26. Victor Zhong, Caiming Xiong, and Richard Socher. 2018. Global-locally self-attentive dialogue state tracker. arXiv preprint arXiv:1805.09655. — [arXiv:1805.09655](https://arxiv.org/abs/1805.09655) find
  27. Zhaopeng Tu, Zhengdong Lu, Yang Liu, Xiaohua Liu, and Hang Li. 2016. Modeling coverage for neural machine translation. arXiv preprint arXiv:1601.04811. find our card
  28. Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. 2017. Attention is all you need. In Advances in neural information processing systems, pages 5998–6008. find
  29. Tsung-Hsien Wen, Milica Gašić, Nikola Mrkšić, Pei-Hao Su, David Vandyke, and Steve Young. 2015. Semantically conditioned LSTM-based natural language generation for spoken dialogue systems. In Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing, pages 1711–1721,… find
  30. Tsung-Hsien Wen, Yishu Miao, Phil Blunsom, and Steve Young. 2017. Latent intention dialogue models. In Proceedings of the 34th International Conference on Machine Learning-Volume 70, pages 3732– 3741. JMLR. org. find
  31. Jason D Williams, Kavosh Asadi, and Geoffrey Zweig. 2017. Hybrid code networks: practical and efficient end-to-end dialog control with supervised and reinforcement learning. In Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 665–… find
  32. Chien-Sheng Wu, Andrea Madotto, Ehsan Hosseini-Asl, Caiming Xiong, Richard Socher, and Pascale Fung. 2019. Transferable multi-domain state generator for task-oriented dialogue systems. In Proceedings of the 57th Conference of the Association for Computational Linguistics, ACL 2019, Florence, Italy,… find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/wang-2020-multi-domain-dialogue-acts-response/