Knowledge-Bridged Causal Interaction Network for Causal Emotion Entailment

Utterance Leverage (statistics) Bridge (graph theory) Causal reasoning
DOI: 10.1609/aaai.v37i11.26641 Publication Date: 2023-06-27T18:11:51Z
ABSTRACT
Causal Emotion Entailment aims to identify causal utterances that are responsible for the target utterance with a non-neutral emotion in conversations. Previous works limited thorough understanding of conversational context and accurate reasoning cause. To this end, we propose Knowledge-Bridged Interaction Network (KBCIN) commonsense knowledge (CSK) leveraged as three bridges. Specifically, construct graph each conversation leverage event-centered CSK semantics-level bridge (S-bridge) capture deep inter-utterance dependencies via CSK-Enhanced Graph Attention module. Moreover, social-interaction serves emotion-level (E-bridge) action-level (A-bridge) connect candidate one, which provides explicit clues Emotional module Actional reason emotion. Experimental results show our model achieves better performance over most baseline models. Our source code is publicly available at https://github.com/circle-hit/KBCIN.
SUPPLEMENTAL MATERIAL
Coming soon ....
REFERENCES (0)
CITATIONS (13)
EXTERNAL LINKS
PlumX Metrics
RECOMMENDATIONS
FAIR ASSESSMENT
Coming soon ....
JUPYTER LAB
Coming soon ....