AI summary

This paper introduces a large-scale persona-based dialogue dataset, built from REDDIT conversations with over 5 million personas and 700 million dialogues, and shows that training with personas improves end-to-end systems, achieving state-of-the-art results on the PERSONA-CHAT dataset via transfer learning.

Abstract

Current dialogue systems are not very 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 of limited size as it 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.