Investigating the effectiveness of data mining techniques in identifying early signs of mental health issues from social media usage
Nature Journal of Emerging Sciences Technologies and Innovations · 29.04.2026 · 7 (4) · 362-376
DOI: 10.65752/6mghk619 ↗Kurz gefasst
Fachartikel in der Zeitschrift Nature Journal of Emerging Sciences Technologies and Innovations, online erschienen am 29.04.2026. Verfasst von Rebecca Nneka Oguzor und 2 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
The proliferation of social media has created extensive textual data, offering valuable insights for early detection of mental health conditions. This study investigates the effectiveness of machine learning and deep learning models for classifying mental health posts. A structured pipeline was implemented, encompassing text preprocessing, stratified dataset splitting, feature extraction, and rigorous model evaluation using Precision, Recall, F1-Score, and ROC-AUC. Four models: Logistic Regression, Support Vector Machine (SVM), Random Forest, and a fine-tuned BERTwere systematically compared. The BERT model consistently outperformed traditional approaches, achieving a Precision of 0.81, Recall of 0.83, F1-Score of 0.82, and ROC-AUC of 0.90. Random Forest followed with an F1-Score of 0.77 and ROC-AUC of 0.84, SVM achieved an F1-Score of 0.72 and ROC-AUC of 0.80, while Logistic Regression showed the lowest performance (F1-Score 0.71, ROC-AUC 0.77). Confusion matrix and curve analyses confirmed BERT’s superior ability to distinguish between positive and negative posts while maintaining a strong balance between precision and recall. These results highlight the advantage of transformer-based architectures in capturing semantic and contextual nuances in social media text. Despite dataset and computational limitations, the proposed framework provides a robust, reproducible, and high-performing approach for automated mental health detection and early intervention.
Zitieren
APA
Oguzor, R. N., Oriko, B. C., & Okoro, F. N. (2026). Investigating the effectiveness of data mining techniques in identifying early signs of mental health issues from social media usage. Nature Journal of Emerging Sciences Technologies and Innovations, 7(4), 362-376. https://doi.org/10.65752/6mghk619
Vancouver
Oguzor RN, Oriko BC, Okoro FN. Investigating the effectiveness of data mining techniques in identifying early signs of mental health issues from social media usage. Nature Journal of Emerging Sciences Technologies and Innovations. 2026;7(4):362-376. doi:10.65752/6mghk619
BibTeX
@article{Oguzor2026,
author = {Oguzor, Rebecca Nneka and Oriko, Blessing Chimaobi and Okoro, Favour Ngozi},
title = {Investigating the effectiveness of data mining techniques in identifying early signs of mental health issues from social media usage},
journal = {Nature Journal of Emerging Sciences Technologies and Innovations},
year = {2026},
volume = {7},
number = {4},
pages = {362--376},
doi = {10.65752/6mghk619}
}
EUVIMED