Empathetic Response Generation via Emotion Cause Transition Graph

Qian, Yushan; Wang, Bo; Lin, Ting-En; Zheng, Yinhe; Zhu, Ying; Zhao, Dongming; Hou, Yuexian; Wu, Yu-Chuan; Li, Yongbin

doi:10.1109/icassp49357.2023.10095652

Cited by 5 publications

(1 citation statement)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, empathy includes both affective and cognitive empathy [3]. So, some studies have focused on cognitive empathy recently, such as exploring the emotional cause of the context [11,17,26], predicting dialog act in the conversation [1,34], introducing external knowledge bases for reasoning to help generate more informative and empathetic responses [15,19,30]. However, in current empathetic dialogue systems, most of the work uses the dialogue context to predict the emotion and the dialog act in the next dialog turn or introduces external knowledge bases for reasoning without dynamically taking the user feedback into account.…”

Section: Related Workmentioning

confidence: 99%

Grafting Fine-Tuning and Reinforcement Learning for Empathetic Emotion Elicitation in Dialog Generation

Zhu,

Wang,

Zhao

et al. 2023

Frontiers in Artificial Intelligence and Applications

View full text Add to dashboard Cite

For human-like dialogue systems, it is significant to inject the empathetic ability or elicit the opposite’s positive emotions, while existing studies mostly only focus on either of the above two research lines. In this work, we propose a novel and grafted task named Empathetic Emotion Elicitation Dialog to make a dialog system able to possess both aspects of ability simultaneously. We do not train an empathetic dialog system and an emotion elicitation dialog system separately and then simply concatenate the responses generated by these two systems, which will cause illogical and repetitive responses. Instead, we propose a unified solution: (1) To generate empathetic responses and emotion elicitation responses within the same semantic space, we design a unified framework. (2) The unified framework has three stages which first retrieve the empathetic and emotion elicitation exemplars as external knowledge, then fine-tune the emotion/action prediction on a pre-trained language model to enhance the empathetic ability, and finally model the user feedback by reinforcement learning to enhance the emotion elicitation ability. Experiments show that our method outperforms the baselines in the response generation quality and simultaneously empathizes with the user and elicits their positive emotions.

show abstract