Customer-Review Sentiment Analysis Web App

This project is a web app that uses sentiment analysis to enable users get the most recent positive and negative reviews for any business by putting in the name and location.

• Web App link : here
• Docker hub repository : here

Features to be added:

• direct link to instagram pages.
• direct link to websites.
• choose number of reviews.

Table of Contents

• Features
• Architecture
  • Backend API
  • Frontend
• Getting Started
  • Prerequisites
  • Setup Instructions
    • Backend
    • Frontend
• API Endpoints
  • GET /
  • POST /get_reviews
• Key Technologies
• Deployment
  • Backend (Azure Web App Service)
  • Frontend (Netlify)
• How It Works
• Improvements and Roadmap
• Contributing
• License

Features

• Recent Reviews: Focuses on the latest reviews for timely insights.
• Sentiment Analysis: Utilizes Natural Language Processing (NLP) to determine whether reviews are positive, negative, or neutral.
• Interactive Interface: Fully responsive web interface for seamless user experience.

Architecture

The application is divided into two primary components:

1. Backend API

• Built using FastAPI.
• Incorporates Selenium for web scraping customer reviews from Google.
• Uses NLTK Vader for sentiment analysis.
• Packaged in a Docker container (hosted on Docker Hub) with Chrome and ChromeDriver pre-installed for web scraping functionality.
• Hosted on Azure Web App Service.

2. Frontend

• Developed with React (JavaScript, HTML, and CSS).
• Hosted on Netlify for a fast, reliable, and scalable frontend experience.

---

Getting Started

Prerequisites

Ensure the following tools are installed:

• Docker
• Python 3.9+
• Node.js (for frontend development)
• Google Chrome (optional for debugging Selenium locally)

Setup Instructions

Backend

1. Clone the repository:

   git clone https://github.com/0layiw0la/Customer-Review.git
   cd customer-review/API

2. Build and run the Docker container:

   docker build -t customer-review-api .
   docker run -p 8000:8000 customer-review-api

   OR

3. Pull Docker image

   docker pull 0lay1w0la/customerreviewsapi

4. Run Docker container

   docker run -p 8000:8000 customer-review-api

Frontend

1. Navigate to the Frontend directory:

   cd ../Frontend

2. Install dependencies:

   npm install

3. Start the development server:

   npm run dev

---

API Endpoints

Api Link here

GET /

• Description: Test CORS setup.
• Response:

  { "message": "cors works" }

POST /get_reviews

• Description: Scrapes reviews for a given business and performs sentiment analysis.
• Request Body:

  {  
    "business_name": "Example Business",  
    "location_name": "City, State"  
  }

• Response:

  {  
    "positive_reviews": { "review1": "date1", "review2": "date2" },  
    "negative_reviews": { "review3": "date3", "review4": "date4" },  
    "avg_rating": 4.5  
  }

---

Key Technologies

• FastAPI: For API development.
• Selenium: For web scraping reviews from Google.
• NLTK Vader: For sentiment polarity scoring.
• Docker: For containerization and easy deployment.
• React: For the frontend user interface.

---

Deployment

Backend (Azure Web App Service)

• The API is containerized and deployed using Azure's Web App Service.
• The Docker image is hosted publicly on Docker Hub. here

Frontend (Netlify)

• The frontend is deployed on Netlify, ensuring fast and scalable delivery of the app.

---

How It Works

1. User Input: Enter a business name and location.
2. Web Scraping: Backend scrapes Google reviews using Selenium.
3. Sentiment Analysis: Reviews are analyzed using NLTK Vader to classify them as positive, negative, or neutral.
4. Results: Displays the average rating, recent positive and negative reviews, and their dates.

---

Improvements and Roadmap

• Expand review sources beyond Google.
• Add advanced filtering and categorization of reviews.
• Support for multiple languages.
• Upgrade hosting plan for more instances and faster response times.

---

Contributing

Contributions are welcome! Please fork the repository, make changes, and submit a pull request.

---

License

This project is licensed under the MIT License.