Research Article
Advancing Mental Health Support: An Automated Classification System using Text and Audio Inputs to Identify Depression and Suicidal Tendencies
@INPROCEEDINGS{10.4108/eai.28-4-2025.2357774, author={Sumanth Talasila and Grace Shalini T and Shravan Chowdary and Geetha P}, title={Advancing Mental Health Support: An Automated Classification System using Text and Audio Inputs to Identify Depression and Suicidal Tendencies }, proceedings={Proceedings of the 4th International Conference on Information Technology, Civil Innovation, Science, and Management, ICITSM 2025, 28-29 April 2025, Tiruchengode, Tamil Nadu, India, Part I}, publisher={EAI}, proceedings_a={ICITSM PART I}, year={2025}, month={10}, keywords={mental health depression suicidal tendencies automated classification text analysis audio analysis machine learning logistic regression real-time feedback mental health support multimodal input tailored resources}, doi={10.4108/eai.28-4-2025.2357774} }- Sumanth Talasila
Grace Shalini T
Shravan Chowdary
Geetha P
Year: 2025
Advancing Mental Health Support: An Automated Classification System using Text and Audio Inputs to Identify Depression and Suicidal Tendencies
ICITSM PART I
EAI
DOI: 10.4108/eai.28-4-2025.2357774
Abstract
This work builds a contemporary automatic method for assessing depression and suicidality from written narratives and speech. The machine learning algorithms, namely Logistic Regression, Decision Tree, Random Forest, and Multinomial Naive Bayes, are used for accurate mental state classification of users. The online evaluation is based on Logistic Regression, which is the best performing model of those tested for both accuracy and reliability of results. With this, the platform offers quick response coupled with special video suggestions to help users manage their mental health. In contrast to questionnaires and those entering face-to-face consultation situation, this system provides a flexible service and easy access mental health support system to users who require the instant intervention. The platform is an important resource for mental health evaluation and treatment because of the utilization of technology inputs, feedback in real-time, as well as tailor made contents. It uses data anonymization methods to guarantee the privacy of the user, as well as secure functions. The project highlights the importance of early detection and intervention of mental health status, and supports global mental health campaigns. It is hoped that by making it immediately available, such a system will help to break down social barriers to help-seeking and community mental health.
