E-CORE: Emotion Correlation Enhanced Empathetic Dialogue Generation
- URL: http://arxiv.org/abs/2311.15016v1
- Date: Sat, 25 Nov 2023 12:47:39 GMT
- Title: E-CORE: Emotion Correlation Enhanced Empathetic Dialogue Generation
- Authors: Fengyi Fu, Lei Zhang, Quan Wang, Zhendong Mao
- Abstract summary: We propose a novel emotion correlation enhanced empathetic dialogue generation framework.
Specifically, a multi-resolution emotion graph is devised to capture context-based emotion interactions.
We then propose an emotion correlation enhanced decoder, with a novel correlation-aware aggregation and soft/hard strategy.
- Score: 33.57399405783864
- License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
- Abstract: Achieving empathy is a crucial step toward humanized dialogue systems.
Current approaches for empathetic dialogue generation mainly perceive an
emotional label to generate an empathetic response conditioned on it, which
simply treat emotions independently, but ignore the intrinsic emotion
correlation in dialogues, resulting in inaccurate emotion perception and
unsuitable response generation. In this paper, we propose a novel emotion
correlation enhanced empathetic dialogue generation framework, which
comprehensively realizes emotion correlation learning, utilization, and
supervising. Specifically, a multi-resolution emotion graph is devised to
capture context-based emotion interactions from different resolutions, further
modeling emotion correlation. Then we propose an emotion correlation enhanced
decoder, with a novel correlation-aware aggregation and soft/hard strategy,
respectively improving the emotion perception and response generation.
Experimental results on the benchmark dataset demonstrate the superiority of
our model in both empathetic perception and expression.
Related papers
Err
This list is automatically generated from the titles and abstracts of the papers in this site.
This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.