AI-Based Mental Health Chatbot
Collaborators: Dr. S.N. Omkar (Supervisor), Dhruv Shinde (Collaborator)
Institution: Computer Intelligence Lab, Indian Institute of Science (IISc)
Project Overview
The project aimed to develop an AI-powered chatbot grounded in Cognitive Behavioral Therapy (CBT) principles to support college students dealing with mental health issues. Built on the Rasa framework, the chatbot integrates psychological techniques with conversational AI to provide scalable, personalized mental health support.
The chatbot was designed to guide users through structured interventions inspired by CBT, offering practical tools for managing emotions and thoughts effectively.
Responsibilities and Contributions
-
Literature Review:
-
Conducted a comprehensive review of existing research on mental health chatbots, CBT principles, and conversational AI.
-
Identified gaps in current solutions to inform design and functionality.
-
-
Questionnaire Design:
-
Developed a detailed list of questions for chatbot trials, focusing on usability, engagement, and therapeutic efficacy.
-
Ensured questions aligned with CBT techniques to encourage meaningful interactions with users.
-
-
Research on Randomized Control Trials (RCTs):
-
Explored methodologies for implementing RCTs to evaluate chatbot effectiveness.
-
Proposed using college students with mental health challenges as the target group for trials.
-
-
Analysis of Existing Solutions:
-
Evaluated competing mental health chatbots to understand their models, strengths, and limitations.
-
Analyzed conversational flows, response accuracy, and therapeutic outcomes.
-
-
Bot Evaluation and Modifications:
-
Reviewed the initial prototype and provided actionable recommendations for improvement.
-
Suggested modifications to enhance user experience, align with CBT methodologies, and ensure accessibility for college students.
-
Key Learnings and Impact
-
Multidisciplinary Integration: Gained experience in blending psychological principles with cutting-edge AI technology to create a meaningful tool for mental health support.
-
Research and Evaluation Skills: Strengthened skills in literature review, experimental design, and competitor analysis.
-
Social Relevance: Contributed to a project aimed at addressing a critical need for scalable mental health solutions, particularly among young adults.
Outcome and Future Directions
The chatbot prototype was optimized based on the research and evaluations conducted. Future work will involve implementing RCTs to validate its effectiveness, scaling its deployment to a larger audience, and exploring additional psychological frameworks to enhance functionality.