Leveraging DistilBERT for Predictive Analytics and Insights in Mental Health Disorder Classification

G. Angelpriya, J. Nirmala Gandhi, V. Venkatesh Guru

2025

Abstract

Mental health disorders, especially depression, are a major global challenge and therefore early detection is required for timely intervention. Conventional assessment forms can be inaccessible, subjective or delayed; hence the demand for more scalable and automated solutions. In this work, we utilize recent developments within the field of natural language processing (NLP) to create a reliable system for classifying depressive and non-depressive language with DistilBERT, a distilled transformer model. In response to these issues, a synthetic dataset was carefully designed to capture a range of emotional expressions that are sensitive and diverse while ensuring ethical considerations. From the dataset generation and pre-processing methods through model training and evaluation, it reaches a 100% accuracy on the test set. A user-friendly, accessible and responsive-interface deployed prediction framework for real-time inference of inputs using the trained model. It is created to serve as a mental health monitor rather than a diagnostic replacement, placing ethics at the forefront. As the field of NLP matures, this research seeks to connect state‐of‐the‐art models and real physical world problems when it comes to early detection and intervention for mental health conditions with potential of affecting a large set of people in an ethically sound way.

Paper Citation

in Harvard Style

Angelpriya G., Gandhi J. and Guru V. (2025). Leveraging DistilBERT for Predictive Analytics and Insights in Mental Health Disorder Classification. In Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - ICRDICCT`25; ISBN 978-989-758-777-1, SciTePress, pages 811-816. DOI: 10.5220/0013906100004919

in Bibtex Style

@conference{icrdicct`2525,
author={G. Angelpriya and J. Gandhi and V. Guru},
title={Leveraging DistilBERT for Predictive Analytics and Insights in Mental Health Disorder Classification},
booktitle={Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - ICRDICCT`25},
year={2025},
pages={811-816},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013906100004919},
isbn={978-989-758-777-1},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - ICRDICCT`25
TI - Leveraging DistilBERT for Predictive Analytics and Insights in Mental Health Disorder Classification
SN - 978-989-758-777-1
AU - Angelpriya G.
AU - Gandhi J.
AU - Guru V.
PY - 2025
SP - 811
EP - 816
DO - 10.5220/0013906100004919
PB - SciTePress