🧠 Student Support AI (Advanced)

A Context-Aware Mental Health Assistant built with Python & Scikit-Learn. Live Demo: Click Here to Open App

🚀 Project Overview

This application helps students manage stress by analyzing their raw text input. Unlike simple sentiment analyzers, this AI uses a Double-Brain Architecture to understand both Emotion and Topic simultaneously.

🛠️ How it Works

The system runs two Logistic Regression models in parallel:

1. Emotion Classifier: Detects 6 emotions (Joy, Sadness, Anger, Fear, Love, Surprise). Trained on 16,000 data points.
2. Topic Classifier: Detects the root cause (Academic, Relationship, Career, Mental Health).

Based on the combination (e.g., Academic + Fear), the system looks up a specialized psychological strategy to assist the user.

💻 Tech Stack

• Language: Python
• ML Libraries: Scikit-Learn, Pandas, Joblib
• NLP: TF-IDF Vectorization, Text Preprocessing
• Web Framework: Streamlit
• Hosting: Streamlit Cloud

📊 Key Features

• Real-time NLP: Instantly classifies text input.
• Balanced Training: Implemented Class Weights to handle imbalanced datasets (fixing the "Surprise" class recall).
• Personalized Logic: Maps (Emotion + Topic) pairs to specific actionable advice.

✍️ Author

Upamada Ekanayake - AI & Data Science Student