FastAPI Recommendation Service

This project implements a recommendation service using FastAPI and PostgreSQL. Docker Containers are added for ruture scalability. The service provides recommendations for posts based on user behavior such as likes and dislikes.

Features

• Recommendation System: Utilizes a CatBoostClassifier model to predict user preferences and recommend posts accordingly.
• Database Integration: Integrates with a PostgreSQL database using SQLAlchemy for data storage and retrieval.
• Asynchronous Support: Leverages asynchronous programming to handle concurrent requests efficiently.
• Chunked Data Retrieval: Retrieves large datasets from the database in chunks to optimize performance.
• Parquet File Handling: Loads data from Parquet files if available, else loads from the database.
• Logging: Implements logging using the Loguru library for detailed runtime information.
• Docker Containerization: Utilizes Docker for containerizing the PostgreSQL database and FastAPI service for easy deployment and scaling.
• Configurable: Configuration settings are loaded from a YAML file for easy customization.

Installation

Installation

1. Clone the repository:

   git clone https://github.com/RadionNazmiev/recommendation_system_project.git

2. Set up the database environment variables in the .env file:

   PG_HOST=postgres_db
   PG_PORT=5432
   PG_DB=your_database_name
   PG_USER=your_database_user
   PG_PASS=your_database_password
   API_PORT=8000

3. Build and run the Docker containers using Docker Compose:

   docker-compose up --build

4. Access the FastAPI service at http://localhost:8000.

Usage

• GET /post/recommendations: Get recommended posts for a user at a specific time.
• GET /posts/{post_id}: Get details of a specific post by ID.
• GET /posts: Get a list of posts with optional limit parameter.
• GET /users/{user_id}: Get details of a specific user by ID.
• GET /users: Get a list of users with optional limit parameter.
• GET /feed: Get the feed, optionally filtered by user_id or post_id with a limit.

Contributing

Contributions are welcome! Please check the CONTRIBUTING.md file for guidelines.

License

This project is licensed under the MIT License. See the LICENSE file for details.