What type of study is this?

This is a Quantitative Study study (also classified as: Experimental Study).

September 12, 2025MathematicsOpen Access

RAMHA: A Hybrid Social Text-Based Transformer with Adapter for Mental Health Emotion Classification

View Full Paper

Ask AI

Bookmark

Authors

Hybrid deep learning model ramha detects mental health emotions in social media, indicating effective early intervention strategies.

RAMHA achieves 92% accuracy in binary emotion classification on the GoEmotions dataset, and shows high performance across multiple metrics.
In extensive experiments, RAMHA consistently outperformed eight baseline models, including BERT and LSTM, highlighting its superiority in precision and recall.
The model integrates RoBERTa and adapter layers, optimized with focal loss to tackle class imbalance in emotion classification tasks.
The findings suggest that RAMHA could enhance emotion-aware mental health screening in both clinical and social environments.

Kumar et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark