Spider Game Recommender

An AI-powered game recommendation system that leverages collaborative filtering and K-Nearest Neighbors (KNN) to provide personalized game suggestions based on user behavior and preferences.

Overview

The Spider Game Recommender is a comprehensive solution that combines:

• A Flask API serving AI models for game recommendations
• A C# backend API for handling user interactions and data management
• An Angular frontend for a seamless user experience

Features

• Personalized game recommendations using collaborative filtering and content-based methods
• Real-time API for dynamic suggestions
• Integration with Steam for game images
• User profile management and favorite game tracking
• Dockerized deployment for easy scaling and consistency

Technical Stack

• AI/ML: Python, scikit-learn, pandas
• Backend: Flask (Python), ASP.NET Core (C#)
• Frontend: Angular
• Deployment: Docker, Contabo Linux Ubuntu server

Key Components

1. Data Processing Pipeline
   • Data ingestion from multiple sources
   • Feature extraction using TF-IDF vectorization
   • KNN model training for content-based filtering
2. Recommendation Engine
   • Collaborative filtering for user-based recommendations
   • Content-based filtering using game features
3. APIs
   • Flask API for serving ML models
   • C# API for user management and application logic
4. Web Application
   • Angular-based frontend for user interactions
   • Display of game recommendations with images

Deployment

The system is deployed on a Contabo Linux Ubuntu server using Docker containers for both the Flask and C# APIs.

API Documentation

API documentation is available through Swagger UI. Key endpoints include:

• GET /api/Home/recommendations: Fetch game recommendations
• PUT /api/Home/update-favorite/{id}: Update user's favorite games
• POST /api/Home/get-collaborative-filter-recommends: Get recommendations using collaborative filtering

Future Improvements

• Implement model retraining capabilities
• Enhance error handling and input validation
• Implement logging for performance monitoring
• Explore more complex models or ensemble methods for improved accuracy