Elamur Research⌕
☰
home / papers / 2016 / Multi-level Gated Recurrent Neural Network for dialog act classification

Multi-level Gated Recurrent Neural Network for dialog act classification

Summary

The paper addresses dialog act (DA) classification: assigning an intention label to each utterance in a conversation. This is hard because many utterances in conversational data are extremely short, such as "Okay" or a laughter marker, and earlier neural models mostly ignored surrounding turns. Li and Wu propose a hierarchical, two-level gated recurrent neural network (GRNN). The lower level reads the words of each utterance and max-pools the GRU outputs into a sentence vector. A small feed-forward network encodes non-textual cues, such as utterance length and the utterance's position in the dialog. The two vectors are concatenated and passed to an upper-level GRNN, which runs across consecutive utterances, so the model can choose which context to keep or forget. Experiments on the Switchboard Dialog Act (SWDA) corpus show that this design beats prior state-of-the-art systems by more than 6 accuracy points. The approach needs little hand-crafted feature engineering. It also shows that modelling conversation context is the main driver of better DA tagging, especially for short, ambiguous turns.

Key points

  • Data: SWDA uses 43 DAMSL-based tags. 41.4% of its utterances have fewer than 5 words, about 61% have fewer than 10, and 3,253 consist of a single word.
  • Best result: The full model (multi-level GRNN+non-textual) reaches 79.37% accuracy, up from the previous best of 73.1%. That is a gain of 6.27 points over Lee and Dernoncourt (2016) and 6.77 points over Shen and Lee (2016).
  • Baselines: CNN scores 68.25, single-level GRNN 69.75 and non-textual FFNN alone 43.60.
  • Effect of non-textual features: Adding them to a single model gives about 2–2.6 points.
  • Effect of context: The upper-level GRNN over utterances matters most. Placing it on top of the CNN yields 77.14, roughly 9 points above the plain CNN, and multi-level GRNN alone yields 77.65.
  • Error analysis: Context lets the final model correctly tag ambiguous short turns like "Okay" as bk (Response Acknowledgement), where the single-level GRNN fails.

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

Cited by

8 papers of our library cite it (of the 140673 whose bibliographies we read); 2 of them have a page here, 6 not yet.

Cite

The paper itself, not this page. BibTeX from the library's record of the work.

@inproceedings{li2016multi,
  title = {{Multi-level Gated Recurrent Neural Network for dialog act classification}},
  author = {Li, Wei and Wu, Yunfang},
  booktitle = {Proceedings of COLING 2016, the 26th International Conference on Computational Linguistics: Technical Papers},
  year = {2016},
  url = {https://aclanthology.org/C16-1185/}
}

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. Martı́n Abadi, Ashish Agarwal, Paul Barham, Eugene Brevdo, Zhifeng Chen, Craig Citro, Greg S. Corrado, Andy Davis, Jeffrey Dean, Matthieu Devin, Sanjay Ghemawat, Ian Goodfellow, Andrew Harp, Geoffrey Irving, Michael Isard, Yangqing Jia, Rafal Jozefowicz, Lukasz Kaiser, Manjunath Kudlur, Josh Levenberg, Dan Mané, Rajat Monga, Sherry Moore, Derek Murray, Chris Olah, Mike Schuster, Jonathon Shlens, Benoit Steiner, Ilya Sutskever, Kunal Talwar, Paul Tucker, Vincent Vanhoucke, Vijay Vasudevan, Fernanda Viégas, Oriol Vinyals, Pete Warden, Martin Wattenberg, Martin Wicke, Yuan Yu, and Xiaoqiang Zheng. 2015. TensorFlow: Large-scale machine learning on heterogeneous systems. Software available from tensorflow.org. find
  2. James Allen and Mark Core. 1997. Draft of damsl: Dialog act markup in several layers. Unpublished manuscript, 2. find
  3. John Langshaw Austin and JO Urmson. 1962. How to Do Things with Words. The William James Lectures Delivered at Harvard University in 1955.[Edited by James O. Urmson.]. Clarendon Press. find
  4. Yoshua Bengio, Patrice Simard, and Paolo Frasconi. 1994. Learning long-term dependencies with gradient descent is difficult. Neural Networks, IEEE Transactions on, 5(2):157–166. ↗
  5. Harry Bunt, Jan Alexandersson, Jae-Woong Choe, Alex Chengyu Fang, Koiti Hasida, Volha Petukhova, Andrei Popescu-Belis, and David R Traum. 2012. Iso 24617-2: A semantically-based standard for dialogue annotation. In LREC, pages 430–437. Citeseer. find our card
  6. Kyunghyun Cho, Bart Van Merriënboer, Caglar Gulcehre, Dzmitry Bahdanau, Fethi Bougares, Holger Schwenk, and Yoshua Bengio. 2014. Learning phrase representations using rnn encoder-decoder for statistical machine translation. arXiv preprint arXiv:1406.1078. ↗
  7. Junyoung Chung, Caglar Gulcehre, KyungHyun Cho, and Yoshua Bengio. 2014. Empirical evaluation of gated recurrent neural networks on sequence modeling. arXiv preprint arXiv:1412.3555. ↗
  8. Ronan Collobert and Jason Weston. 2007. Fast semantic extraction using a novel neural network architecture. In Annual meeting-association for computational linguistics, volume 45, page 560. find
  9. Ronan Collobert and Jason Weston. 2008. A unified architecture for natural language processing: Deep neural networks with multitask learning. In Proceedings of the 25th international conference on Machine learning, pages 160–167. ACM. find
  10. Ronan Collobert, Jason Weston, Léon Bottou, Michael Karlen, Koray Kavukcuoglu, and Pavel Kuksa. 2011. Natural language processing (almost) from scratch. The Journal of Machine Learning Research, 12:2493–2537. find our card
  11. Rajdip Dhillon, Sonali Bhagat, Hannah Carvey, and Elizabeth Shriberg. 2004. Meeting recorder project: Dialog act labeling guide. Technical report, DTIC Document. ↗
  12. Jeffrey L Elman. 1990. Finding structure in time. Cognitive science, 14(2):179–211. ↗
  13. John J Godfrey, Edward C Holliman, and Jane McDaniel. 1992. Switchboard: Telephone speech corpus for research and development. In Acoustics, Speech, and Signal Processing, 1992. ICASSP-92., 1992 IEEE International Conference on, volume 1, pages 517–520. IEEE. find
  14. Sepp Hochreiter and Jürgen Schmidhuber. 1997. Long short-term memory. Neural computation, 9(8):1735–1780. ↗
  15. Haifeng Hu, Bingquan Liu, Baoxun Wang, Ming Liu, and Xiaolong Wang. 2013. Multimodal dbn for predicting high-quality answers in cqa portals. find
  16. Ozan Irsoy and Claire Cardie. 2014. Opinion mining with deep recurrent neural networks. In EMNLP, pages 720–728. find
  17. Dan Jurafsky, Elizabeth Shriberg, and Debra Biasca. 1997. Switchboard swbd-damsl shallow-discourse-function annotation coders manual. Institute of Cognitive Science Technical Report, pages 97–102. find
  18. Nal Kalchbrenner, Edward Grefenstette, and Phil Blunsom. 2014. A convolutional neural network for modelling sentences. arXiv preprint arXiv:1404.2188. ↗ our card
  19. Su Nam Kim, Lawrence Cavedon, and Timothy Baldwin. 2010. Classifying dialogue acts in one-on-one live chats. In Proceedings of the 2010 Conference on Empirical Methods in Natural Language Processing, pages 862–871. Association for Computational Linguistics. find our card
  20. Yoon Kim. 2014. Convolutional neural networks for sentence classification. arXiv preprint arXiv:1408.5882. Diederik Kingma and Jimmy Ba. arXiv:1412.6980. 2014. Adam: A method for stochastic optimization. arXiv preprint Ji Young Lee and Franck Dernoncourt. 2016. Sequential short-text classification with recurrent and convolutional neural networks. arXiv preprint arXiv:1603.03827. ↗ our card
  21. Max M Louwerse and Scott A Crossley. 2006. Dialog act classification using n-gram algorithms. In FLAIRS Conference, pages 758–763. find
  22. Tomas Mikolov, Martin Karafiát, Lukas Burget, Jan Cernockỳ, and Sanjeev Khudanpur. 2010. Recurrent neural network based language model. In INTERSPEECH, volume 2, page 3. find
  23. Tomas Mikolov, Kai Chen, Greg Corrado, and Jeffrey Dean. 2013. Efficient estimation of word representations in vector space. arXiv preprint arXiv:1301.3781. ↗
  24. Hamid Palangi, Li Deng, Yelong Shen, and Jianfeng Gao. 2015. Deep sentence embedding using long short-term memory networks: Analysis and application to information retrieval. IEEE/ACM Transactions on Audio Speech & Language Processing, 24(4):694–707. ↗
  25. Norbert Reithinger and Martin Klesen. 1997. Dialogue act classification using language models. In EuroSpeech. Citeseer. find
  26. John R Searle. 1969. Speech acts: An essay in the philosophy of language, volume 626. Cambridge university press. find
  27. Sheng-syun Shen and Hung-yi Lee. 2016. Neural attention models for sequence classification: Analysis and application to key term extraction and dialogue act detection. arXiv preprint arXiv:1604.00077. ↗
  28. Joao Silva, Luı́sa Coheur, Ana Cristina Mendes, and Andreas Wichert. 2011. From symbolic to sub-symbolic information in question classification. Artificial Intelligence Review, 35(2):137–154. ↗
  29. Andreas Stolcke, Noah Coccaro, Rebecca Bates, Paul Taylor, Carol Van Ess-Dykema, Klaus Ries, Elizabeth Shriberg, Daniel Jurafsky, Rachel Martin, and Marie Meteer. 2000. Dialogue act modeling for automatic tagging and recognition of conversational speech. Computational linguistics, 26(3):339–373. ↗ our card
  30. Dinoj Surendran and Gina-Anne Levow. 2006. Dialog act tagging with support vector machines and hidden markov models. In INTERSPEECH. find
  31. Kaisheng Yao, Baolin Peng, Yu Zhang, Dong Yu, Geoffrey Zweig, and Yangyang Shi. 2014. Spoken language understanding using long short-term memory neural networks. In Spoken Language Technology Workshop (SLT), 2014 IEEE, pages 189–194. IEEE. find
  32. Yucan Zhou, Qinghua Hu, Jie Liu, and Yuan Jia. 2015. Combining heterogeneous deep neural networks with conditional random fields for chinese dialogue act recognition. Neurocomputing, 168:408–417. ↗

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/papers/li-2016-multi-level-gated-recurrent-neural/