Multi-Task Learning and Adapted Knowledge Models for Emotion-Cause Extraction

TL;DR

This paper introduces a joint multi-task learning approach that integrates common-sense knowledge models to improve emotion recognition and cause detection in text, advancing fine-grained emotion analysis.

Contribution

It presents novel methods combining common-sense knowledge with multi-task learning for joint emotion and cause detection, which enhances performance over previous approaches.

Findings

Performance improved with common-sense reasoning

Joint modeling benefits both tasks

Thorough analysis provides insights into model behavior

Abstract

Detecting what emotions are expressed in text is a well-studied problem in natural language processing. However, research on finer grained emotion analysis such as what causes an emotion is still in its infancy. We present solutions that tackle both emotion recognition and emotion cause detection in a joint fashion. Considering that common-sense knowledge plays an important role in understanding implicitly expressed emotions and the reasons for those emotions, we propose novel methods that combine common-sense knowledge via adapted knowledge models with multi-task learning to perform joint emotion classification and emotion cause tagging. We show performance improvement on both tasks when including common-sense reasoning and a multitask framework. We provide a thorough analysis to gain insights into model performance.