Add correlation explanation and improve health recommendation NLU

feat: Add correlation explanation action for stroke risk factors; update NLU and rules for new intent

feat: Enhance NLU intents and examples for greetings, farewells, gratitude, and apologies; add stroke risk tips rule

refactor: Simplify rules and stories by removing overwrite flows and enhancing chitchat interactions

feat: Update domain responses and stories for improved user interaction and data overwrite confirmation

- Updated `docker-compose.yml` to include the `docs` directory for proper access to static assets.
- Fixed banner loading in `home.py` by correcting file path and handling exceptions.
- Moved `pages` directory to `web` for consistency and updated imports in `app.py`.
- Improved data preprocessing in `data_preprocessor.py` by implementing outlier removal using Z-score.
- Enhanced recommendation predictions in `recommendations.py` by displaying prediction probabilities.
- Added sidebar sections in `app.py` to dynamically display available models and evaluations.
- Refactored file structure for better organization and maintainability.

- Added Z-score based outlier detection and removal in data_preprocessor.py.
- Restructured src/app.py into a modular format with separate pages for Home, Data Analysis, Recommendations, and Chatbot.
- Enhanced error handling and session management within the Streamlit application.
- Updated project directory to include new modules and maintain code readability.