INTELLIGENT MENTAL HEALTH PREDICTION SYSTEM USING MACHINE LEARNING, ENSEMBLE TECHNIQUES, AND LARGE LANGUAGE MODELS

Abstract

The increasing prevalence of mental health disorders highlights the urgent need for intelligent systems capable of early detection and prevention. With the rapid growth of social media platforms, users often express emotional states, psychological stress, and behavioral changes through their digital footprints. This research proposes an Intelligent Mental Health Prediction System that integrates Machine Learning (ML), Ensemble Learning, and Large Language Models (LLMs) to detect and predict potential mental disorders from social media data. The proposed system begins with data collection and preprocessing from multiple social media sources, focusing on text, sentiment, and behavioral cues indicative of mental stress, depression, or anxiety. Machine learning algorithms such as Support Vector Machines (SVM), Random Forests, and Gradient Boosting are trained to identify early signs of psychological distress. An ensemble framework combines the strengths of these classifiers to improve robustness, reduce bias, and enhance overall predictive accuracy. In parallel, Large Language Models (e.g., GPT, BERT) are utilized for contextual understanding, emotion recognition, and semantic feature extraction, enabling deeper insight into user sentiment and intent. The hybrid integration of ML and LLM-based representations allows the system to handle linguistic nuances, slang, and context-specific variations prevalent in online communication. Experimental evaluation on benchmark mental health datasets demonstrates that the proposed framework achieves higher accuracy, precision, and recall compared to conventional single-model approaches. Furthermore, the system offers interpretability through explainable AI techniques, allowing mental health professionals to understand underlying triggers and patterns. By combining data-driven analytics with advanced language intelligence, this system provides a proactive approach to digital mental health monitoring, facilitating early intervention and personalized psychological support. It sets a foundation for responsible AI applications in social media–based mental health prediction, bridging the gap between technology and human well-being.