An AI-Assisted Mood Tracking and Mental Health Alert System Using Sentiment Analysis and Facial Emotion Recognition

Mental health conditions such as anxiety, depression, and emotional instability are rapidly increasing worldwide, yet mental well-being continues to be neglected due to stigma, lack of awareness, and insufficient access to professional help. A significant limitation in existing mood tracking applications is their reliance on either manual mood input or single-modality analysis, resulting in inaccurate and inconsistent mental state interpretation. This research proposes an AI-assisted multi-modal mood tracking and mental health alert system that integrates text-based sentiment analysis, facial emotion detection, and conversational AI to deliver continuous and automated psychological assessment. The system is implemented using Python and Streamlit as the UI layer, supported by SQLite as a lightweight secure database, and incorporates VADER sentiment analysis, DeepFace for real-time emotion recognition, and a Gemini-powered mental health chatbot with fallback rule-based responses. The platform allows users to maintain digital journals, track mood trends, analyze emotional patterns through graphs and word clouds, and receive real-time wellness guidance. Experimental results demonstrate the feasibility of multimodal mood detection and automated emotional analytics, providing an efficient and accessible approach for early mental health awareness and intervention.