Exploiting Emotion-Semantic Correlations for Empathetic Response Generation

arXiv:2402.17437v1 Announce Type: new
Abstract: Empathetic response generation aims to generate empathetic responses by understanding the speaker's emotional feelings from the language of dialogue. Recent methods capture emotional words in the language of communicators and construct them as static vectors to perceive nuanced emotions. However, linguistic research has shown that emotional words in language are dynamic and have correlations with other grammar semantic roles, i.e., words with semantic meanings, in grammar. Previous methods overlook these two characteristics, which easily lead …

abstract arxiv construct correlations cs.ai cs.cl dialogue emotion emotions feelings generate language research responses semantic speaker them type understanding vectors words