Title: Chatbot for mental health diagnosis using data augmentation techniques and deep learning

Authors: Neel Ghoshal; Vaibhav Bhartia; B.K. Tripathy; A. Tripathy

Addresses: SCOPE, Vellore Institute of Technology, Vellore, 632014, Tamil Nadu, India ' SCOPE, Vellore Institute of Technology, Vellore, 632014, Tamil Nadu, India ' SCORE, Vellore Institute of Technology, Vellore, 632014, Tamil Nadu, India ' Carnegie Mellon University, 5000 Forbes Ave, Pittsburgh, PA 15213, USA

Abstract: Mental illness has become widespread among people world over. Although several chatbot based models have been designed, their efficient utilisation by people is not properly confirmed. In this paper a customised chatbot framework is proposed and developed using natural language understanding (NLU) mechanisms, which comprises a unique two-tier modular functionality of an empathetic conversational model with a simultaneous implementation of a classification model. The framework provides a holistic service to a user. The dataset is prepared manually to include the various mental health diseases and the appropriate responses provided by professionals. The model uses conversational therapeutic data and uses RASA as its structural framework, which is trained to perform efficient and vicarious dialogue with a user. The mental health-based categorical dataset undergoes various models such as logistic regression, decision tree random forest, naive Bayes and Google BERT leading to an accuracy of 91.08%.

Keywords: chatbot; mental health; natural language processing; NLP; deep learning; soft computing.

DOI: 10.1504/IJES.2024.144352

International Journal of Embedded Systems, 2024 Vol.17 No.3/4, pp.171 - 182

Received: 29 Apr 2023
Accepted: 16 Sep 2023
Published online: 10 Feb 2025 *