Amazon-Recommendation-System is a Windows-friendly project for exploring how product recommendation works with Amazon rating data. It uses common recommendation methods to compare results and show how each model performs on the same data.

This project includes:

• Rank-based recommendations
• User-to-user filtering
• Item-to-item filtering
• SVD-based recommendations
• Basic evaluation
• Visual data analysis

It is built for the MIT Applied AI & Data Science Program Elective Project and uses plain data science methods that help you understand what a recommender system does.

Use this link to visit the project page:

Visit the GitHub project page

On Windows, open the page, then:

1. Click the green or blue Code button.
2. Choose Download ZIP.
3. Save the file to your computer.
4. Right-click the ZIP file and choose Extract All.
5. Open the extracted folder.
6. Look for the main project files and run the notebook or Python script based on the included setup.

If the project includes Jupyter Notebook files, open them with Jupyter Notebook or JupyterLab. If it includes Python scripts, run them from a Python setup with the needed packages installed.

Before you start, make sure your Windows PC has:

• Windows 10 or later
• At least 4 GB of RAM
• Enough free disk space for the dataset and project files
• Internet access for downloading the repository
• Python 3.9 or later
• Jupyter Notebook or JupyterLab if the project uses notebooks

If you want smoother performance with larger data files, 8 GB of RAM or more helps.

This project centers on Amazon product ratings data and shows how different recommendation methods compare.

You can expect files for:

• Data loading
• Data cleaning
• Model building
• Model testing
• Plotting and charts
• Notebook-based analysis

The project topics include:

• Amazon dataset
• Collaborative filtering
• Data analysis
• Data science
• EDA
• Hyperparameter tuning
• KNN
• Machine learning
• Recommendation systems
• Recommender system

Go to the GitHub page and download the ZIP file.

Right-click the ZIP file and choose Extract All.

Open the extracted folder in File Explorer.

If Python is not already on your PC:

1. Download Python from the official Python website.
2. Install it.
3. Check the box that says Add Python to PATH.

Press Windows + R, type cmd, and press Enter.

If the project uses Python scripts or notebooks, install the common packages with:

• pip install pandas
• pip install numpy
• pip install matplotlib
• pip install scikit-learn
• pip install jupyter

If the repository includes a requirements.txt file, use:

• pip install -r requirements.txt

If the project uses notebooks:

• Open Command Prompt in the project folder
• Run jupyter notebook
• Open the notebook file in your browser

If the project uses scripts:

• Run the main Python file from the project folder
• Follow the file names and instructions in the repository

This project compares several recommender system methods.

This model gives a simple list of popular items. It works well as a first step when you want a basic recommendation set.

This method finds users with similar rating patterns. It then suggests items that similar users liked.

This method looks at items that have similar rating patterns. It then recommends items that match what a user already liked.

This model uses matrix factorization to find hidden patterns in the rating data. It helps produce more refined recommendations.

The project does more than generate recommendations. It also checks how well each method works.

Common checks include:

• Model comparison
• Accuracy review
• Pattern detection
• Rating distribution analysis
• Visual charts for easier review

This helps you see which method gives the best result for the dataset.

After you open the notebook or script:

1. Load the dataset files.
2. Run the data cleaning steps.
3. Build the recommendation model.
4. Review the output lists.
5. Compare the results across models.
6. Look at the charts and evaluation output.

If the project asks for a user ID, product ID, or item name, enter the value shown in the example cells or input prompts.

You may see folders and files such as:

• data — raw or cleaned ratings data
• notebooks — analysis notebooks
• src — Python code files
• models — saved model output
• plots or images — charts and graphs
• README files — project notes and setup help

If the folder names are different, use the files that match the same purpose.

This project is useful if you want to:

• Study how recommender systems work
• Compare simple and advanced recommendation methods
• Explore Amazon rating data
• Learn basic machine learning on real data
• Review data analysis and evaluation steps

Check that Python is installed and added to PATH.

Install it again with:

• pip install jupyter

Then run:

• jupyter notebook

Install the package named in the error message with pip install package-name.

Make sure you extracted the ZIP file before opening the project.

Open the project from the main folder so the code can find the data files.

• Repository Name: Amazon-Recommendation-System
• Description: MIT Applied AI & Data Science Program Elective Project. End-to-end recommendation system project using Amazon product ratings data, implementing rank-based, collaborative filtering (User-User, Item-Item), and SVD models, with evaluation and visual analysis.
• Topics: amazon-dataset, collaborative-filtering, data-analysis, data-science, eda, hyperparameter-tuning, knn, machine-learning, recommendation-systems, recommender-system

Download or visit the project page