Recommender System Website

A full-stack e-commerce recommendation system website built with React and Node.js, featuring personalized product recommendations based on user reviews and interactions.

🚀 Features

• User Authentication: Sign up and login system with unique user ID generation
• Product Browsing: Browse and search through smartphone products
• Personalized Recommendations: AI-powered product recommendations based on user preferences
• Product Reviews: Users can add and view product reviews
• Shopping Cart: Add products to cart and manage quantities
• Order Management: Place orders and track order history
• Product Likes: Like/favorite products for personalized recommendations
• User Profile: Manage user information including name, age, gender, and address

🛠️ Tech Stack

Frontend

• React - UI library
• Vite - Build tool and dev server
• React Router - Client-side routing
• Axios - HTTP client
• Context API - State management

Backend

• Node.js - Runtime environment
• Express.js - Web framework
• MongoDB - Database
• Mongoose - MongoDB object modeling
• CORS - Cross-origin resource sharing

📋 Prerequisites

• Node.js (v14 or higher)
• npm or yarn
• MongoDB database (local or Atlas)

🔧 Installation

1. Clone the repository

git clone <repository-url>
cd RecommenderSystem_Website

2. Backend Setup

cd backend
npm install

Create a .env file in the backend directory:

MONGO_URI=your_mongodb_connection_string
PORT=5001
FRONTEND_URL=http://localhost:5173
NODE_ENV=development

3. Frontend Setup

cd frontend
npm install

Create a .env file in the frontend directory (optional for local development):

VITE_API_URL=http://localhost:5001/api

Or:

VITE_BACKEND_URL=http://localhost:5001

4. Seed the Database (Optional)

To populate the database with sample data:

cd backend
npm run seed

🚀 Running the Application

Development Mode

1. Start the backend server:

cd backend
npm run dev

The backend will run on http://localhost:5001

2. Start the frontend (in a new terminal):

cd frontend
npm run dev

The frontend will run on http://localhost:5173

Production Build

Frontend:

cd frontend
npm run build

Backend:

cd backend
npm start

🔐 Environment Variables

Backend (.env)

MONGO_URI=mongodb://localhost:27017/recommender_db
PORT=5001
FRONTEND_URL=http://localhost:5173
NODE_ENV=development

Frontend (.env)

VITE_API_URL=http://localhost:5001/api
# OR
VITE_BACKEND_URL=http://localhost:5001

🚢 Deployment

The application is deployed on:

• Frontend: Vercel
• Backend: Render

For detailed deployment instructions, see DEPLOYMENT.md

📝 Key Features Explained

User Signup

• Users can sign up with reviewer name, age, gender, and address
• System automatically generates a unique 26-character user ID
• Users are automatically logged in after signup

Recommendations

• Recommendations are based on user reviews and aspect-based sentiment analysis
• System uses cosine similarity to match user preferences with products
• Recommendations consider product ratings and popularity

Model Retraining

The system includes a function to check if the recommendation model needs retraining:

• Triggers when total reviews reach 2,500 OR
• Triggers when total interactions (likes + orders) reach 3,000