Predicting Mental Health Disorders based on Sentiment Analysis

The increasing prevalence of mental health disorders such as anxiety, depression, and stress calls for innovative, data-driven approaches to early detection and intervention. This research investigates the use of sentiment analysis to predict mental health disorders from social media text, utilizing machine learning techniques to enhance the accuracy and interpretability of mental health classification. We developed and evaluated multiple machine learning models, including Random Forest, Extra Trees, AdaBoost, and Logistic Regression, using a curated dataset of user-generated content reflecting various mental health states. Our rigorous experimentation demonstrates the potential of sentiment analysis combined with advanced machine learning models in identifying mental health conditions. The study highlights that while Random Forest and Extra Trees excel in prediction accuracy, Logistic Regression offers a balanced trade-off between performance and model stability, making it suitable for real-world applications. These findings contribute to the growing field of AI-driven mental health assessment, offering a foundation for integrating multimodal data, model interpretability, and ethical implementation.