Training Millions of Personalized Dialogue Agents

Pierre-Emmanuel Mazaré, Samuel Humeau, Martin Raison, Antoine Bordes


Abstract
Current dialogue systems fail at being engaging for users, especially when trained end-to-end without relying on proactive reengaging scripted strategies. Zhang et al. (2018) showed that the engagement level of end-to-end dialogue models increases when conditioning them on text personas providing some personalized back-story to the model. However, the dataset used in Zhang et al. (2018) is synthetic and only contains around 1k different personas. In this paper we introduce a new dataset providing 5 million personas and 700 million persona-based dialogues. Our experiments show that, at this scale, training using personas still improves the performance of end-to-end systems. In addition, we show that other tasks benefit from the wide coverage of our dataset by fine-tuning our model on the data from Zhang et al. (2018) and achieving state-of-the-art results.
Anthology ID:
D18-1298
Volume:
Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing
Month:
October-November
Year:
2018
Address:
Brussels, Belgium
Venue:
EMNLP
SIG:
SIGDAT
Publisher:
Association for Computational Linguistics
Note:
Pages:
2775–2779
Language:
URL:
https://aclanthology.org/D18-1298
DOI:
10.18653/v1/D18-1298
Bibkey:
Cite (ACL):
Pierre-Emmanuel Mazaré, Samuel Humeau, Martin Raison, and Antoine Bordes. 2018. Training Millions of Personalized Dialogue Agents. In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, pages 2775–2779, Brussels, Belgium. Association for Computational Linguistics.
Cite (Informal):
Training Millions of Personalized Dialogue Agents (Mazaré et al., EMNLP 2018)
Copy Citation:
PDF:
https://preview.aclanthology.org/update-css-js/D18-1298.pdf
Video:
 https://vimeo.com/305943582