Linguistic Signatures for Enhanced Emotion Detection

TL;DR

This paper investigates how linguistic features can serve as interpretable signals for emotion detection in text, demonstrating that incorporating these features into transformer models improves performance and robustness across multiple datasets.

Contribution

It introduces a method to extract emotion-specific linguistic signatures and shows that enriching transformer models with these features enhances emotion recognition accuracy.

Findings

Enriched models achieve up to +2.4 macro F1 on GoEmotions.

Linguistic features complement neural representations for better robustness.

Linguistic signatures are consistent across diverse datasets.

Abstract

Emotion detection is a central problem in NLP, with recent progress driven by transformer-based models trained on established datasets. However, little is known about the linguistic regularities that characterize how emotions are expressed across different corpora and labels. This study examines whether linguistic features can serve as reliable interpretable signals for emotion recognition in text. We extract emotion-specific linguistic signatures from 13 English datasets and evaluate how incorporating these features into transformer models impacts performance. Our RoBERTa-based models enriched with high level linguistic features achieve consistent performance gains of up to +2.4 macro F1 on the GoEmotions benchmark, showing that explicit lexical cues can complement neural representations and improve robustness in predicting emotion categories.