Emotion-Cause Pair Extraction: A New Task to Emotion Analysis in Texts

TL;DR

This paper introduces the emotion-cause pair extraction (ECPE) task, enabling simultaneous extraction of emotions and their causes in texts, overcoming limitations of previous methods that required prior emotion annotation.

Contribution

It proposes a novel ECPE task and a two-step multi-task learning approach for joint emotion and cause extraction and pairing in documents.

Findings

ECPE is feasible and effective on benchmark data

The approach outperforms existing methods in emotion-cause extraction

Joint extraction improves accuracy and applicability in real-world scenarios

Abstract

Emotion cause extraction (ECE), the task aimed at extracting the potential causes behind certain emotions in text, has gained much attention in recent years due to its wide applications. However, it suffers from two shortcomings: 1) the emotion must be annotated before cause extraction in ECE, which greatly limits its applications in real-world scenarios; 2) the way to first annotate emotion and then extract the cause ignores the fact that they are mutually indicative. In this work, we propose a new task: emotion-cause pair extraction (ECPE), which aims to extract the potential pairs of emotions and corresponding causes in a document. We propose a 2-step approach to address this new ECPE task, which first performs individual emotion extraction and cause extraction via multi-task learning, and then conduct emotion-cause pairing and filtering. The experimental results on a benchmark emotion cause corpus prove the feasibility of the ECPE task as well as the effectiveness of our approach.