📌 Recommendation System

Welcome to the Recommendation System repository! 🚀 This project provides personalized recommendations based on user interactions and media attributes.

📖 Overview

This system suggests relevant media based on:

• Content Similarity (categories, keywords, authors)
• Collaborative Filtering (user preferences, likes, and views)

⚙️ Features

✅ One-Hot Encoding for categories, keywords, and authors ✅ Cosine similarity for content-based recommendations ✅ User similarity analysis for collaborative filtering ✅ Media correlation using a pivot table ✅ Hybrid recommendation combining content & collaborative methods

📌 Usage

To get recommendations, use the getRecommand function:

recommended_items = getRecommand(data, likes, views, media_id)

This function returns a dictionary of recommended media items based on hybrid filtering.

📊 How It Works

1. Content-Based Filtering 📚

   • Extracts categories, keywords, and authors
   • Applies One-Hot Encoding
   • Calculates cosine similarity between items

2. Collaborative Filtering 👥

   • Analyzes user likes & views
   • Creates a user-item matrix
   • Computes correlations for recommendation

3. Hybrid Approach 🔀

   • Combines both methods
   • Prioritizes top-ranked recommendations
   • Shuffles recommendations for diversity

🤝 Contributing

Contributions are welcome! Feel free to open an issue or submit a pull request. 😊

Happy coding! 🚀