GitHub - doncoriolan/google-fake-reviews: Check Reviews and verify Authenticity

2 min read Original article ↗

Google Reviews AI Checker

This project is a system for detecting Google fake reviews using a TensorFlow AI model. It consists of a Python backend that handles data processing and AI prediction, and a Chrome extension frontend that extracts review data from webpages and displays the analysis.

Tutorial

tutorial.mp4

Overview

The system works as follows:

  1. Chrome Extension: Extracts the HTML of a webpage containing reviews.
  2. Flask Backend: Receives the HTML, processes it to extract key review data points (reviewer name, review count, star rating, etc.).
  3. AI Model: A TensorFlow model scores the authenticity of each review based on the extracted data.
  4. Results: The backend sends an analysis summary back to the Chrome extension, which displays it to the user.

Components

Chrome Extension (/chrome_extension)

The frontend is a browser extension that provides the user interface. For detailed information on setup and usage, see the Chrome Extension readme.

Python Backend (/python)

The backend is a Flask application that contains the data processing, AI model, and API endpoint. For detailed information on the backend architecture, model training, and setup, see the Python Backend readme.

Future Improvements

  • Language detection for lazy reviews
  • Remove manual need for scrolling

License

This project is released into the public domain under the Unlicense.