Supervised Prototypical Contrastive Learning for Emotion Recognition in Conversation

TL;DR

This paper introduces a novel supervised contrastive learning approach using prototypical networks for emotion recognition in conversations, effectively handling class imbalance and improving performance on standard benchmarks.

Contribution

It proposes the SPCL loss with curriculum learning for ERC, addressing class imbalance and sample difficulty without large batch requirements, achieving state-of-the-art results.

Findings

State-of-the-art performance on three benchmarks

Effective handling of class imbalance

Demonstrated benefits of curriculum learning

Abstract

Capturing emotions within a conversation plays an essential role in modern dialogue systems. However, the weak correlation between emotions and semantics brings many challenges to emotion recognition in conversation (ERC). Even semantically similar utterances, the emotion may vary drastically depending on contexts or speakers. In this paper, we propose a Supervised Prototypical Contrastive Learning (SPCL) loss for the ERC task. Leveraging the Prototypical Network, the SPCL targets at solving the imbalanced classification problem through contrastive learning and does not require a large batch size. Meanwhile, we design a difficulty measure function based on the distance between classes and introduce curriculum learning to alleviate the impact of extreme samples. We achieve state-of-the-art results on three widely used benchmarks. Further, we conduct analytical experiments to demonstrate the effectiveness of our proposed SPCL and curriculum learning strategy. We release the code at https://github.com/caskcsg/SPCL.