COSMIC: COmmonSense knowledge for eMotion Identification in Conversations

TL;DR

COSMIC is a novel framework that leverages commonsense knowledge like mental states and causal relations to improve emotion recognition in conversations, addressing key challenges and achieving state-of-the-art results.

Contribution

It introduces a new commonsense-based approach for emotion recognition in conversations, enhancing context understanding and emotion shift detection.

Findings

Achieves new state-of-the-art results on four datasets.

Effectively models interactions using mental states, events, and causal relations.

Addresses challenges in context propagation and emotion differentiation.

Abstract

In this paper, we address the task of utterance level emotion recognition in conversations using commonsense knowledge. We propose COSMIC, a new framework that incorporates different elements of commonsense such as mental states, events, and causal relations, and build upon them to learn interactions between interlocutors participating in a conversation. Current state-of-the-art methods often encounter difficulties in context propagation, emotion shift detection, and differentiating between related emotion classes. By learning distinct commonsense representations, COSMIC addresses these challenges and achieves new state-of-the-art results for emotion recognition on four different benchmark conversational datasets. Our code is available at https://github.com/declare-lab/conv-emotion.