This project aims to get a polarity and subjectivity score of any Yelp (website) product review.

Most of us think should we buy a product or not or maybe should we hire a particular electrician or not? This project helps you out by getting a sentiment analysis of all the reviews of that particular service on Yelp and applies NLP techniques to provide you a score of polarity and subjectivity. But a layman cannot infer scores of such a prediction and here comes Gemini API that uses the score to provide a layman terms understanding of the scores and a suggesting action of whether you should buy it or not. Also using a database you can see the top 5 services with descending order of polarity scores that other users have viewed.

Tech stacks used:

• Python
• Flask
• Flask WTForms
• SQLAlchemy
• Gemini API
• NLTK
• Core ML libraries

The code oprates in 3 primary modules:

• It web scrapes all reviews of a particular product on Yelp.
• It uses Machine Learning model provided by TextBlob to provide inference on the sentiment metrics.
• Finally it passes the sentiment metrics to Google Gemini API to get a detailed inference.

Link: https://sentimento-vvpx.onrender.com/
The project is operating stably but improvements and suggestions are welcome.