DeepEmo: Learning and Enriching Pattern-Based Emotion Representations

TL;DR

DeepEmo introduces a graph-based approach to extract and enrich emotion patterns from text, significantly improving emotion recognition accuracy by leveraging pattern analysis and word embeddings.

Contribution

It presents a novel graph-based method for extracting and enriching emotion patterns, enhancing emotion recognition performance over existing techniques.

Findings

Outperforms most state-of-the-art emotion recognition methods

Demonstrates the effectiveness of pattern enrichment with word embeddings

Provides insights into emotion-oriented pattern properties

Abstract

We propose a graph-based mechanism to extract rich-emotion bearing patterns, which fosters a deeper analysis of online emotional expressions, from a corpus. The patterns are then enriched with word embeddings and evaluated through several emotion recognition tasks. Moreover, we conduct analysis on the emotion-oriented patterns to demonstrate its applicability and to explore its properties. Our experimental results demonstrate that the proposed techniques outperform most state-of-the-art emotion recognition techniques.