Welcome to Amazonia! This is a web-based application that allows Amazon customers to quickly view concise product reviews. The application is built using the Flask and Node.js framework.

1. Clone the repository
2. Install the dependencies

Python

cd python
pip install -r requirements.txt

Javascript

cd javascript
npm install

3. Run the application Python

cd python
python pyserver.py

Javascript

cd javascript
npm run start

For the Chrome extension, please visit this guide for installation instructions: https://support.google.com/chrome_webstore/answer/2664769?hl=en

Visit any Amazon product page, and click the Amazonia extension icon. Then, click "Analyze" to start the analysis

Welcome page

Review Analysis

I used a mixture of Javascript scripts, HTML and CSS to create the page. The extension is built using the Chrome Extension API. The extension is able to communicate with the Node.js server using the Fetch API.

I used Node.js and Puppeteer.js to scrape the Amazon product page, which is then sent to the Flask backend for sentiment analysis and summarization.

I used the OpenAI framework to send requests to the GPT-3 API for summarization. For the sentiment anslysis, I used this Huggingface model: https://huggingface.co/nlptown/bert-base-multilingual-uncased-sentiment