What question did this study set out to answer?

The aim is to improve detection of multiple overlapping emotions in text, especially for low-resource languages.

February 10, 2026Scientific ReportsOpen Access

A hybrid stacked ensemble learning framework for multilabel text emotion detection

View Full Paper

Ask AI

Bookmark

The aim is to improve detection of multiple overlapping emotions in text, especially for low-resource languages.
Implemented a hybrid stacked ensemble framework called Hyb-Stack.
Integrated predictions from four transformer models (BERT, DistilBERT, RoBERTa, mBERT).
Used a Random Forest meta-classifier for enhanced accuracy.
Evaluated on three datasets, including high and low-resource languages.
mBERT achieved the highest F1-scores across datasets: 82.28 on Hausa, 85.33 on Bahasa Indonesia, and 88.90 on SemEval-2018 English.
The EM-9 ensemble set improved performance, exceeding individual models and standard ensemble strategies with F1-scores of 89.48, 88.19, and 90.67.
Overall, the framework displayed effectiveness in multi-label emotion detection across diverse linguistic contexts.

No takes yet. Share an insight, caveat, or question.

Demonstrates enhanced multilabel emotion classification in low-resource languages, suggesting improved sentiment analysis for diverse contexts.

Adamu et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark

Demonstrates enhanced multilabel emotion classification in low-resource languages, suggesting improved sentiment analysis for diverse contexts.