A Weakly Supervised Dataset of Fine-Grained Emotions in Portuguese

TL;DR

This paper presents a weakly supervised, lexical-based dataset for fine-grained emotion recognition in Portuguese, demonstrating its effectiveness by fine-tuning a BERT model with promising results in a low-resource setting.

Contribution

It introduces a novel weakly supervised dataset for fine-grained emotion recognition in Portuguese, suitable for low-resource NLP environments.

Findings

F1-score of 0.64 on validation set

Lexical-based weak supervision is effective for low-resource languages

Dataset enables initial emotion recognition research in Portuguese

Abstract

Affective Computing is the study of how computers can recognize, interpret and simulate human affects. Sentiment Analysis is a common task inNLP related to this topic, but it focuses only on emotion valence (positive, negative, neutral). An emerging approach in NLP is Emotion Recognition, which relies on fined-grained classification. This research describes an approach to create a lexical-based weakly supervised corpus for fine-grained emotion in Portuguese. We evaluated our dataset by fine-tuning a transformer-based language model (BERT) and validating it on a Gold Standard annotated validation set. Our results (F1-score=.64) suggest lexical-based weak supervision as an appropriate strategy for initial work in low resourced environment.