Elamur Research⌕
☰
home / papers / 2018 / Sequicity: Simplifying Task-oriented Dialogue Systems with Single Sequence-to-Sequence Architectures

Sequicity: Simplifying Task-oriented Dialogue Systems with Single Sequence-to-Sequence Architectures

Summary

Conventional task-oriented dialogue systems chain separate modules, and the belief tracker is the weakest link. It typically uses one classifier per slot, which grows heavy as slots and values multiply. It also cannot recognise slot values unseen in training, depends on delexicalisation, and must be pre-trained apart from the rest of the system. Lei et al. replace this pipeline with Sequicity, a framework built on a single seq2seq model. Its central idea is the belief span (bspan): a short text segment, with <Inf> and <Req> fields, that the model writes out at each turn to record constraints and requested slots. Under a Markov assumption, the model reads only the previous bspan, the previous system reply and the current user utterance. It then decodes in two steps: first the new bspan, which drives the knowledge base lookup, then the reply, conditioned on the bspan and the lookup result. The authors build this as Two Stage CopyNet (TSCP) and train it with supervised learning plus optional reinforcement learning. On CamRes676 and KVRET, TSCP beats pipeline baselines in both task success and language quality. It is also roughly ten times smaller and faster to train, and it still works when slot values are out-of-vocabulary.

Key points

  • Sequicity frames dialogue state tracking and response generation as one sequence-generation problem, so the whole system is trained end to end with no separately pre-trained tracker.
  • TSCP extends CopyNet (Gu et al., 2016), since words in bspans and replies are often copied from earlier utterances. Reinforcement learning fine-tuning raises both BLEU and success F1 at modest extra training cost.
  • Evaluation uses CamRes676 and KVRET. In OOV stress tests, the pipeline baseline NDM collapses, while TSCP keeps a reasonable entity match rate because it learns sentence patterns rather than fixed value labels.
  • On KVRET, the authors grow the slot inventory through 79, 144 and 284 distinct slot values by adding domains. TSCP's parameter count stays about an order of magnitude below NDM's and is far less sensitive to this growth.
  • Source code is released at github.com/WING-NUS/sequicity. A simple post-processing script helps handle users who change their minds, but only when slot values are predefined.

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: Natural language processing

Cited by

106 papers of our library cite it (of the 140884 whose bibliographies we read); 6 of them have a page here, 100 not yet.

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/p18-1133

@inproceedings{lei2018sequicity,
  title = {{Sequicity: Simplifying Task-oriented Dialogue Systems with Single Sequence-to-Sequence Architectures}},
  author = {Lei, Wenqiang and Jin, Xisen and Kan, Min-Yen and Ren, Zhaochun and He, Xiangnan and Yin, Dawei},
  booktitle = {Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year = {2018},
  doi = {10.18653/v1/p18-1133},
  url = {https://doi.org/10.18653/v1/p18-1133}
}

References

36 references, 7 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Hongshen Chen, Xiaorui Liu, Dawei Yin, and Jiliang Tang. 2017. A survey on dialogue systems: Recent advances and new frontiers. arXiv preprint arXiv:1711.01731 . — [arXiv:1711.01731](https://arxiv.org/abs/1711.01731) find
  2. Heriberto Cuayáhuitl, Simon Keizer, and Oliver Lemon. 2015. Strategic dialogue management via deep reinforcement learning. arXiv preprint arXiv:1511.08099 . — [arXiv:1511.08099](https://arxiv.org/abs/1511.08099) find
  3. Bhuwan Dhingra, Lihong Li, Xiujun Li, Jianfeng Gao, Yun-Nung Chen, Faisal Ahmed, and Li Deng. 2017. Towards end-to-end reinforcement learning of dialogue agents for information access. In Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers).… find our card
  4. Mihail Eric and Christopher D Manning. 2017a. A copy-augmented sequence-to-sequence architecture gives good performance on task-oriented dialogue. find
  5. Mihail Eric and Christopher D Manning. 2017b. Key-value retrieval networks for task-oriented dialogue. SIGDIAL . find
  6. Jonas Gehring, Michael Auli, David Grangier, and Yann N Dauphin. 2017. A convolutional encoder model for neural machine translation. ACL . find our card
  7. Jiatao Gu, Zhengdong Lu, Hang Li, and Victor OK Li. 2016. Incorporating copying mechanism in sequence-to-sequence learning. ACL . find our card
  8. Homa B Hashemi, Amir Asiaee, and Reiner Kraft. 2016. Query intent detection using convolutional neural networks. In International Conference on Web Search and Data Mining, Workshop on Query Understanding. find
  9. Matthew Henderson, Blaise Thomson, and Jason D Williams. 2014a. The second dialog state tracking challenge. In SIGDIAL Conference. pages 263–272. find
  10. Matthew Henderson, Blaise Thomson, and Jason D Williams. 2014b. The third dialog state tracking challenge. In Spoken Language Technology Workshop (SLT), 2014 IEEE. IEEE, pages 324–329. find
  11. Matthew Henderson, Blaise Thomson, and Steve Young. 2013. Deep neural network approach for the dialog state tracking challenge. In Proceedings of the SIGDIAL 2013 Conference. pages 467–471. find
  12. Andrej Karpathy and Li Fei-Fei. 2015. Deep visual-semantic alignments for generating image descriptions. In Proceedings of the IEEE conference on computer vision and pattern recognition. pages 3128–3137. find our card
  13. John F Kelley. 1984. An iterative design methodology for user-friendly natural language office information applications. ACM Transactions on Information Systems (TOIS) 2(1):26–41. find
  14. Seokhwan Kim, Luis Fernando DHaro, Rafael E Banchs, Jason D Williams, and Matthew Henderson. 2017. The fourth dialog state tracking challenge. In Dialogues with Social Robots, Springer, pages 435– 449. find
  15. Diederik P Kingma and Jimmy Ba. 2015. Adam: A method for stochastic optimization. find
  16. Sungjin Lee. 2013. Structured discriminative model for dialog state tracking. In SIGDIAL Conference. pages 442–451. find
  17. Jiwei Li, Michel Galley, Chris Brockett, Jianfeng Gao, and Bill Dolan. 2015. A diversity-promoting objective function for neural conversation models. arXiv preprint arXiv:1510.03055 . find our card
  18. Bing Liu and Ian Lane. 2017. Iterative policy learning in end-to-end trainable task-oriented neural dialog models. arXiv preprint arXiv:1709.06136 . — [arXiv:1709.06136](https://arxiv.org/abs/1709.06136) find
  19. Hongyuan Mei, Mohit Bansal, and Matthew R Walter. 2016. What to talk about and how? selective generation using lstms with coarse-to-fine alignment. In NAACL. find our card
  20. Nikola Mrkšić, Diarmuid O Séaghdha, Tsung-Hsien Wen, Blaise Thomson, and Steve Young. 2017. Neural belief tracker: Data-driven dialogue state tracking. ACL . find
  21. Tim Paek and Eric Horvitz. 2000. Conversation as action under uncertainty. In Proceedings of the Sixteenth conference on Uncertainty in artificial intelligence. Morgan Kaufmann Publishers Inc., pages 455–464. find
  22. 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. Association for Computational Linguistics, pages 311–318. find our card
  23. Yangyang Shi, Kaisheng Yao, Le Tian, and Daxin Jiang. 2016. Deep lstm based feature mapping for query classification. In Proceedings of the 2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. pages 1501–1511. find
  24. Joao Silva, Luı́sa Coheur, Ana Cristina Mendes, and Andreas Wichert. 2011. From symbolic to subsymbolic information in question classification. Artificial Intelligence Review 35(2):137–154. find
  25. Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Lukasz Kaiser, and Illia Polosukhin. 2017. Attention is all you need. arXiv preprint arXiv:1706.03762 . find
  26. Tsung-Hsien Wen, Milica Gasic, Nikola Mrksic, Lina M Rojas-Barahona, Pei-Hao Su, Stefan Ultes, David Vandyke, and Steve Young. 2016a. Conditional generation and snapshot learning in neural dialogue systems. EMNLP . find
  27. Tsung-Hsien Wen, Milica Gasic, Nikola Mrkšić, Lina M. Rojas Barahona, Pei-Hao Su, Stefan Ultes, David Vandyke, and Steve Young. 2016b. Conditional generation and snapshot learning in neural dialogue systems. In EMNLP. ACL, Austin, Texas, pages 2153–2162. https://aclweb.org/anthology/D16-1233. ↗
  28. Tsung-Hsien Wen, Yishu Miao, Phil Blunsom, and Steve Young. 2017a. Latent intention dialogue models. ICML . find
  29. Tsung-Hsien Wen, David Vandyke, Nikola Mrksic, Milica Gasic, Lina M Rojas-Barahona, Pei-Hao Su, Stefan Ultes, and Steve Young. 2017b. A network-based end-to-end trainable task-oriented dialogue system. EACL . find
  30. Jason Williams, Antoine Raux, Deepak Ramachandran, and Alan Black. 2013. The dialog state tracking challenge. In Proceedings of the SIGDIAL 2013 Conference. pages 404–413. find
  31. Jason D Williams. 2012. A belief tracking challenge task for spoken dialog systems. In NAACL-HLT Workshop on future directions and needs in the spoken dialog community: tools and data. Association for Computational Linguistics, pages 23–24. find
  32. 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. Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) . find
  33. Steve Young, Milica Gašić, Simon Keizer, François Mairesse, Jost Schatzmann, Blaise Thomson, and Kai Yu. 2010. The hidden information state model: A practical framework for pomdp-based spoken dialogue management. Computer Speech & Language 24(2):150–174. find
  34. Steve Young, Milica Gašić, Blaise Thomson, and Jason D Williams. 2013a. Pomdp-based statistical spoken dialog systems: A review. Proceedings of the IEEE 101(5):1160–1179. find
  35. Steve Young, Milica Gašić, Blaise Thomson, and Jason D Williams. 2013b. Pomdp-based statistical spoken dialog systems: A review. Proceedings of the IEEE 101(5):1160–1179. find
  36. Lantao Yu, Weinan Zhang, Jun Wang, and Yong Yu. 2017. Seqgan: Sequence generative adversarial nets with policy gradient. In AAAI. pages 2852–2858. find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/lei-2018-sequicity-simplifying-task-oriented-dialogue/