CopeMateAI: Personalised Mental Health Counselling Chatbot using Large Language Models (LLMs)

Abstract

"ABSTRACT Globally, mental health issues are abundantly present, but professional psychological services that could be tailored to individual needs are scarce. Main difficulties concerning traditional mental health services include high prices, stigmatisation, and lack of personnel. The increase in prevalence of mental health disorders, especially in today’s society, has led to a demand for personalization and affordability of mental health care seeking. The current existing chatbots are unable to have a personality and interact personally with customers because they cannot understand different personality traits, which makes the communication uninteresting. This paper presents a mental health chatbot which has incorporated the OCEAN personality model, including the Openness, Conscientiousness, Extraversion, Agreeableness, and Neuroticism that are used to customise the interactions with the user. Utilising Gemini 1.5 Pro as its power source, the chatbot can assess user input to identify personality characteristics and then respond with empathy and contextual relevance. This personalised strategy seeks to improve user involvement and the effectiveness of the mental health assistance offered. The design, deployment, and assessment of the chatbot include gathering user interaction data, evaluating personality features, and quantifying the effects on mental well-being. The study shows that the chatbot, Gemini.5 Pro, discriminates against users based on OCEAN personality traits, improving mental health aspects. The Gemini 1.5 Pro model offers more complex conversation and richer responses, making it more effective. The findings suggest that integrating personality trait analysis and advanced language models with the proposed platform could enhance mental health support solutions, enhancing theoretical and practical approaches."