Elamur Research
☰
home / papers / 2016 / A Persona-Based Neural Conversation Model

A Persona-Based Neural Conversation Model

Written for the Research library (claude-opus-5-5) from the paper text, in our own words: the license "arXiv — non-exclusive licence to distribute" does not allow redistributing the text itself. Original — arxiv.org/pdf/1603.06155

Summary

Neural SEQ2SEQ chatbots trained on large human-to-human corpora tend to output the most probable, averaged reply. As a result, one system may say it lives in Los Angeles and then in Madrid, or give different ages to reworded questions. Li et al. address this lack of speaker consistency by giving the generator an explicit persona.

In the Speaker Model, each user has a learned embedding that enters the decoder LSTM at every step, together with the current word vector and the previous hidden state. These embeddings are trained jointly with the rest of the network and need no attribute labels. The Speaker-Addressee Model merges the embeddings of both participants into an interaction vector, so it can capture how people adjust their speech to a specific partner.

Experiments on Twitter conversations and TV series scripts show gains over SEQ2SEQ baselines in BLEU, perplexity and human-judged consistency. The work matters because it shows that background and style can emerge from unlabeled dialogue, which points toward personalized assistants and game characters.

Key points

  • The persona models improve on SEQ2SEQ baselines by up to 20% relative in BLEU and 12% relative in perplexity.
  • In the human consistency evaluation on Twitter, after removing ties, judges rated the Speaker Model at least somewhat more consistent in 56.7% of comparisons. Counting only clear "more consistent" verdicts, it won 6.1% of cases versus 1.6% for the baseline.
  • The Twitter setup uses about 70K speaker embeddings and a 50k-word vocabulary. The motivating inconsistent examples come from a 4-layer SEQ2SEQ model trained on 25 million Twitter snippets.
  • Speaker vectors cluster users by traits such as age or country. This lets the model infer answers a user was never asked, linking London with England and Jakarta with Indonesia. Consistency is still imperfect, for example on age or field of study.
  • The Speaker-Addressee Model is motivated by lexical entrainment. The paper illustrates it with Friends, where Ross speaks differently to Monica than to Rachel.

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: Other

Cited by

308 papers of our library cite it (of the 140673 whose bibliographies we read); 17 of them have a page here, 291 not yet.

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.48550/arxiv.1603.06155 · arXiv 1603.06155

@misc{li2016persona,
  title = {{A Persona-Based Neural Conversation Model}},
  author = {Li, Jiwei and Galley, Michel and Brockett, Chris and Spithourakis, Georgios P. and Gao, Jianfeng and Dolan, Bill},
  year = {2016},
  doi = {10.48550/arxiv.1603.06155},
  eprint = {1603.06155},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.1603.06155}
}

References

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

  1. David Ameixa, Luisa Coheur, Pedro Fialho, and Paulo Quaresma. 2014. Luke, I am your father: dealing with out-of-domain requests by using movies subtitles. In Intelligent Virtual Agents, pages 13–21. Springer. find
  2. Dzmitry Bahdanau, Kyunghyun Cho, and Yoshua Bengio. 2015. Neural machine translation by jointly learning to align and translate. In Proc. of the International Conference on Learning Representations (ICLR). find our card
  3. Rafael E Banchs and Haizhou Li. 2012. IRIS: a chat-oriented dialogue system based on the vector space model. In Proc. of the ACL 2012 System Demonstrations, pages 37–42. find
  4. Yun-Nung Chen, Wei Yu Wang, and Alexander Rudnicky. 2013. An empirical investigation of sparse log-linear models for improved dialogue act classification. In Acoustics, Speech and Signal Processing (ICASSP), 2013 IEEE International Conference on, pages 8317–8321. IEEE. find
  5. Werner Deutsch and Thomas Pechmann. 1982. Social interaction and the development of definite descriptions. Cognition, 11:159–184. find
  6. Michel Galley, Chris Brockett, Alessandro Sordoni, Yangfeng Ji, Michael Auli, Chris Quirk, Margaret Mitchell, Jianfeng Gao, and Bill Dolan. 2015. ∆B LEU : A discriminative metric for generation tasks with intrinsically diverse targets. In Proc. of ACL-IJCNLP, pages 445–450, Beijing, China, July. find
  7. Jianfeng Gao, Xiaodong He, Wen-tau Yih, and Li Deng. 2014. Learning continuous phrase representations for translation modeling. In Proc. of ACL, pages 699–709, Baltimore, Maryland. find
  8. Sepp Hochreiter and Jürgen Schmidhuber. 1997. Long short-term memory. Neural computation, 9(8):1735–1780. find
  9. Alfred Kobsa. 1990. User modeling in dialog systems: Potentials and hazards. AI & society, 4(3):214–231. find
  10. Esther Levin, Roberto Pieraccini, and Wieland Eckert. 2000. A stochastic model of human-machine interaction for learning dialog strategies. IEEE Transactions on Speech and Audio Processing, 8(1):11–23. find
  11. Jiwei Li, Michel Galley, Chris Brockett, Jianfeng Gao, and Bill Dolan. 2016. A diversity-promoting objective function for neural conversation models. In Proc. of NAACL-HLT. find our card
  12. Grace I Lin and Marilyn A Walker. 2011. All the world’s a stage: Learning character models from film. In Proceedings of the Seventh AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment (AIIDE). find
  13. Thang Luong, Ilya Sutskever, Quoc Le, Oriol Vinyals, and Wojciech Zaremba. 2015. Addressing the rare word problem in neural machine translation. In Proc. of ACL, pages 11–19, Beijing, China, July. find our card
  14. Lasguido Nio, Sakriani Sakti, Graham Neubig, Tomoki Toda, Mirna Adriani, and Satoshi Nakamura. 2014. Developing non-goal dialog system based on examples of drama television. In Natural Interaction with Robots, Knowbots and Smartphones, pages 355–361. Springer. find
  15. Franz Josef Och. 2003. Minimum error rate training in statistical machine translation. In Proceedings of the 41st Annual Meeting of the Association for Computational Linguistics, pages 160–167, Sapporo, Japan, July. Association for Computational Linguistics. find
  16. Alice H Oh and Alexander I Rudnicky. 2000. Stochastic language generation for spoken dialogue systems. In Proceedings of the 2000 ANLP/NAACL Workshop on Conversational systems-Volume 3, pages 27–32. find
  17. Kishore Papineni, Salim Roukos, Todd Ward, and Wei-Jing Zhu. 2002. B LEU : a method for automatic evaluation of machine translation. In Proc. of ACL, pages 311–318. find
  18. Roberto Pieraccini, David Suendermann, Krishna Dayanidhi, and Jackson Liscombe. 2009. Are we there yet? research in commercial spoken dialog systems. In Text, Speech and Dialogue, pages 3–13. Springer. find
  19. Adwait Ratnaparkhi. 2002. Trainable approaches to surface natural language generation and their application to conversational dialog systems. Computer Speech & Language, 16(3):435–455. find
  20. Alan Ritter, Colin Cherry, and William B Dolan. 2011. Data-driven response generation in social media. In Proceedings of the Conference on Empirical Methods in Natural Language Processing, pages 583– 593. find
  21. Jost Schatztnann, Matthew N Stuttle, Karl Weilhammer, and Steve Young. 2005. Effects of the user model on simulation-based learning of dialogue strategies. In Automatic Speech Recognition and Understanding, 2005 IEEE Workshop on, pages 220– 225. find
  22. Iulian V Serban, Alessandro Sordoni, Yoshua Bengio, Aaron Courville, and Joelle Pineau. 2015. Building end-to-end dialogue systems using generative hierarchical neural network models. In Proc. of AAAI. find our card
  23. Lifeng Shang, Zhengdong Lu, and Hang Li. 2015. Neural responding machine for short-text conversation. In ACL-IJCNLP, pages 1577–1586. find our card
  24. Alessandro Sordoni, Michel Galley, Michael Auli, Chris Brockett, Yangfeng Ji, Meg Mitchell, Jian-Yun Nie, Jianfeng Gao, and Bill Dolan. 2015. A neural network approach to context-sensitive generation of conversational responses. In Proc. of NAACL-HLT. find our card
  25. Ilya Sutskever, Oriol Vinyals, and Quoc V Le. 2014. Sequence to sequence learning with neural networks. In Advances in neural information processing systems (NIPS), pages 3104–3112. find
  26. Jörg Tiedemann. 2009. News from OPUS – a collection of multilingual parallel corpora with tools and interfaces. In Recent advances in natural language processing, volume 5, pages 237–248. find
  27. Oriol Vinyals and Quoc Le. 2015. A neural conversational model. In Proc. of ICML Deep Learning Workshop. find our card
  28. Wolfgang Wahlster and Alfred Kobsa. 1989. User models in dialog systems. Springer. find
  29. Marilyn A Walker, Rashmi Prasad, and Amanda Stent. 2003. A trainable generator for recommendations in multimodal dialog. In INTERSPEECH. find
  30. Marilyn A Walker, Ricky Grant, Jennifer Sawyer, Grace I Lin, Noah Wardrip-Fruin, and Michael Buell. 2011. Perceived or not perceived: Film character models for expressive nlg. In Interactive Storytelling, pages 109–121. Springer. find
  31. Marilyn A Walker, Grace I Lin, and Jennifer Sawyer. 2012. An annotated corpus of film dialogue for learning and characterizing character style. In LREC, pages 1373–1378. find
  32. William Yang Wang, Ron Artstein, Anton Leuski, and David Traum. 2011. Improving spoken dialogue understanding using phonetic mixture models. In FLAIRS Conference. find
  33. Tsung-Hsien Wen, Milica Gasic, Nikola Mrkšić, Pei-Hao Su, David Vandyke, and Steve Young. 2015. Semantically conditioned LSTM-based natural language generation for spoken dialogue systems. In Proc. of EMNLP, pages 1711–1721, Lisbon, Portugal, September. Association for Computational Linguistics. find
  34. Kaisheng Yao, Geoffrey Zweig, and Baolin Peng. 2015. Attention with intention for a neural network conversation model. CoRR, abs/1510.08565. — [arXiv:1510.08565](https://arxiv.org/abs/1510.08565) find
  35. 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