Personalised E-Commerce Product Recommendation System

Overview

This project aims to create a personalized product recommendation system for e-commerce platforms. It leverages user interaction data, purchase history, and product information to generate tailored recommendations, enhancing the shopping experience and potentially increasing sales.

Features

• Data Exploration and Preprocessing: Jupyter notebooks for exploring, cleaning, and preparing the dataset for the recommendation algorithm.
• Recommendation Algorithm: Implementation of an interest calculation algorithm to score user interest in products based on past interactions and purchases.
• Web Application: A Flask web application to showcase the recommendation system in action, allowing users to browse products, view recommendations, and simulate purchases.

Project Structure

• Code For Recommendation System/: Contains all the Jupyter notebooks and CSV files related to the recommendation algorithm.
  • Data Exploration.ipynb: Notebook for initial data exploration.
  • Data Generation.ipynb: Notebook for generating synthetic data for testing.
  • Interest Calculation-Algorithm.ipynb: Notebook detailing the algorithm for calculating user interest scores.
  • Recommendation.ipynb: Notebook for generating product recommendations.
  • CSV files: Include user.csv, product.csv, purchases.csv, interactions.csv, interest_scores.csv, and all_recommendations.csv for algorithm inputs and outputs.
• Code for Website to showcase Recommendation System/: Contains the Flask application and static assets.
  • app.py: The Flask application.
  • templates/: HTML templates for the web interface.
  • static/: CSS and image files for the web application.
  • CSV files: Same as above, used by the web application to display recommendations.
• Proof of Concept PPT.pdf: A presentation outlining the concept and implementation of the project.
• Report-Personalised Product Recommendation System.pdf: A detailed report on the project, including methodology, results, and conclusions.

Getting Started

To run the recommendation system:

1. Clone the repository.
2. Navigate to the Code For Recommendation System/ directory to explore the Jupyter notebooks.
3. To run the web application:
   • Ensure you have Python and Flask installed.
   • Navigate to the Code for Website to showcase Recommendation System/ directory.
   • Run app.py using Flask.
   • Access the web application at http://localhost:5000.

Technologies Used

• Python: For data processing and the web server.
• Jupyter Notebook: For data exploration and algorithm development.
• Flask: For the web application.
• HTML/CSS: For the web interface.

Contributing

Contributions to improve the recommendation system or the web application are welcome. Please fork the repository, make your changes, and submit a pull request.