This repo contains an in-depth Sentiment analysis on Amazon product reviews with full code explanation. This is a Multi-classification problem solved using NLP/NLU (with nltk, TextBlob, VADER) along with various Machine Learning and Deep Learning models (RoBERTa). Offers also an Amazon ASIN-based review scraper module.
- An Amazon review scraper based on the ASIN given in input;
- A Sentiment Analysis on Amazon product reviews using NLU/NLP techniques along with ML and pre-trained DL models. The analysis is structured over 4 Notebooks: from Data Cleaning to EDA, then with Feature Analysis and Engineering, and finally the Model Selection, Validation and Explainability part.
