Enhancing Cognitive Models of Emotions with Representation Learning

TL;DR

This paper introduces a deep learning framework that generates detailed emotion embeddings, enabling improved emotion classification, hierarchical analysis, and alignment with psychological models like Plutchik's wheel.

Contribution

It presents a novel method combining contextualized embeddings and probing models for fine-grained emotion representation and analysis.

Findings

Achieved state-of-the-art accuracy on emotion classification.

Derived an emotion hierarchy graph from layer analysis.

Generated emotion representations comparable to psychological models.

Abstract

We present a novel deep learning-based framework to generate embedding representations of fine-grained emotions that can be used to computationally describe psychological models of emotions. Our framework integrates a contextualized embedding encoder with a multi-head probing model that enables to interpret dynamically learned representations optimized for an emotion classification task. Our model is evaluated on the Empathetic Dialogue dataset and shows the state-of-the-art result for classifying 32 emotions. Our layer analysis can derive an emotion graph to depict hierarchical relations among the emotions. Our emotion representations can be used to generate an emotion wheel directly comparable to the one from Plutchik's\LN model, and also augment the values of missing emotions in the PAD emotional state model.