Hope Speech Detection in Social Media English Corpora: Performance of Traditional and Transformer Models

TL;DR

This paper compares traditional machine learning models and transformer architectures for hope speech detection in social media, finding transformers generally outperform traditional models in accuracy and semantic understanding.

Contribution

It provides a comprehensive evaluation of traditional and transformer models on hope speech detection, highlighting the superior performance of transformers on small datasets.

Findings

Transformers achieved higher precision and recall than traditional models.

Traditional models like SVM and logistic regression performed well with macro-F1 of 0.78.

Transformers achieved weighted F1 of 0.79 and accuracy of 0.80.

Abstract

The identification of hope speech has become a promised NLP task, considering the need to detect motivational expressions of agency and goal-directed behaviour on social media platforms. This proposal evaluates traditional machine learning models and fine-tuned transformers for a previously split hope speech dataset as train, development and test set. On development test, a linear-kernel SVM and logistic regression both reached a macro-F1 of 0.78; SVM with RBF kernel reached 0.77, and Na\"ive Bayes hit 0.75. Transformer models delivered better results, the best model achieved weighted precision of 0.82, weighted recall of 0.80, weighted F1 of 0.79, macro F1 of 0.79, and 0.80 accuracy. These results suggest that while optimally configured traditional machine learning models remain agile, transformer architectures detect some subtle semantics of hope to achieve higher precision and recall in hope speech detection, suggesting that larges transformers and LLMs could perform better in small datasets.