Emo2Vec: Learning Generalized Emotion Representation by Multi-task Training

TL;DR

Emo2Vec is a generalized emotion embedding learned through multi-task training on six emotion-related tasks, outperforming existing affective representations and achieving competitive results when combined with GloVe.

Contribution

This paper introduces Emo2Vec, a new emotion representation learned via multi-task learning across diverse emotion-related tasks, with improved performance over prior affective embeddings.

Findings

Emo2Vec outperforms Sentiment-Specific Word Embedding and DeepMoji embeddings.

Combining Emo2Vec with GloVe yields competitive results on multiple tasks.

Multi-task training enables effective generalization of emotional semantics.

Abstract

In this paper, we propose Emo2Vec which encodes emotional semantics into vectors. We train Emo2Vec by multi-task learning six different emotion-related tasks, including emotion/sentiment analysis, sarcasm classification, stress detection, abusive language classification, insult detection, and personality recognition. Our evaluation of Emo2Vec shows that it outperforms existing affect-related representations, such as Sentiment-Specific Word Embedding and DeepMoji embeddings with much smaller training corpora. When concatenated with GloVe, Emo2Vec achieves competitive performances to state-of-the-art results on several tasks using a simple logistic regression classifier.