What question did this study set out to answer?

This research aims to improve user engagement in dialogue systems by utilizing a large dataset of personalized personas.

January 1, 2018Open Access

Training Millions of Personalized Dialogue Agents

Puntos clave

This research aims to improve user engagement in dialogue systems by utilizing a large dataset of personalized personas.
Introduced a new dataset with 5 million personas and 700 million persona-based dialogues.
Trained end-to-end dialogue models on this dataset and fine-tuned existing models to enhance performance.
Conducted experiments comparing performance with prior datasets to assess improvements.
Training on personalized dialogue data significantly increased engagement levels in end-to-end models.
Fine-tuning using the new dataset achieved state-of-the-art performance on benchmark tasks.

Resumen

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.

Me gusta

Guardar

Ver artículo completo