📚 Book Recommender System (User-Based Collaborative Filtering)

This project is a simple and clean implementation of a User-Based Book Recommendation System using collaborative filtering and cosine similarity. It is designed for learning purposes and demonstrates the core logic of recommendation engines.

---

✅ Features

• Loads and merges book and rating datasets
• Builds a user-item rating matrix
• Calculates user-user similarity matrix using cosine similarity
• Recommends top N books for a given user
• Modular and well-documented code
• Includes step-by-step learning scripts and final production version
• Ready to deploy and extend with more complex algorithms

---

📂 Project Structure

book-recommender/
│
├── data/                     # Processed CSV files
│   ├── ratings.csv
│   ├── books.csv
│   ├── merged_ratings_books.csv
│   ├── user_item_matrix.csv
│   └── user_similarity_matrix.csv
│
├── learning_steps/           # Educational step-by-step Python scripts
│   ├── step1_load_data.py
│   ├── step2_matrix_processing.py
│   ├── step3_similarity_calculation.py
│   └── step4_recommend_books.py
│
├── final_project/            # Clean and optimized final implementation
│   ├── main.py
│   └── recommender.py
│
├── demo/                     # Sample test script
│   └── demo_recommendation.py
│
└── README.md                 # Project documentation

---

🧠 Recommendation Algorithm

This project uses User-Based Collaborative Filtering and the cosine similarity metric to find similar users. The system then recommends books that similar users have rated highly but the current user hasn’t interacted with.

---

🚀 How to Run

1. Make sure you have Python 3.x and pandas, scikit-learn installed:

pip install pandas scikit-learn

2. Run the scripts in this order:

# Step-by-step (learning)
python learning_steps/step1_load_data.py
python learning_steps/step2_matrix_processing.py
python learning_steps/step3_similarity_calculation.py

# Final version
python final_project/main.py

---

🧪 Example Output

Enter user ID: 276725
Recommended books for user 276725:
- The Hobbit
- To Kill a Mockingbird
- Harry Potter and the Prisoner of Azkaban
...

---

💡 Use Cases & Extensions

• Can be extended to Item-Based Collaborative Filtering
• Can integrate Matrix Factorization or Neural Collaborative Filtering
• Can be deployed as a Flask API or Streamlit app
• Suitable for portfolio, resume, and beginner AI interviews

---

📌 Author

Alireza Ahmadi
AI Developer | Python Enthusiast | Resume-driven Learner
📧 alireza.ahmadi.dehnavi@gmail.com
🌍 Based in Iran, open to remote work and relocation

GitHub is currently my main portfolio due to internet restrictions in Iran.

---

⭐️ Star this project if you find it useful!