Modeling Content-Emotion Duality via Disentanglement for Empathetic Conversation

TL;DR

This paper introduces CEDual, a disentanglement-based framework that models content and emotion duality in dialogues to improve empathetic response generation, achieving state-of-the-art results.

Contribution

The paper proposes a novel disentanglement approach to separately encode content and emotion, enhancing empathetic response quality in dialogue systems.

Findings

CEDual outperforms previous models on EMPATHETICDIALOGUES dataset

Generated responses are more empathetic and contextually appropriate

Achieves state-of-the-art automatic and human evaluation metrics

Abstract

The task of empathetic response generation aims to understand what feelings a speaker expresses on his/her experiences and then reply to the speaker appropriately. To solve the task, it is essential to model the content-emotion duality of a dialogue, which is composed of the content view (i.e., what personal experiences are described) and the emotion view (i.e., the feelings of the speaker on these experiences). To this end, we design a framework to model the Content-Emotion Duality (CEDual) via disentanglement for empathetic response generation. With disentanglement, we encode the dialogue history from both the content and emotion views, and then generate the empathetic response based on the disentangled representations, thereby both the content and emotion information of the dialogue history can be embedded in the generated response. The experiments on the benchmark dataset EMPATHETICDIALOGUES show that the CEDual model achieves state-of-the-art performance on both automatic and human metrics, and it also generates more empathetic responses than previous methods.