Amazon Review Sentiment Analysis & Product Recommender
Goal: Analyze product reviews and recommend items based on sentiment.
Tech Stack: Python, NLP, TextBlob, Sklearn, Recommender Systems
Overview
Classified Amazon reviews by sentiment and recommended related products to improve customer experience and conversion.
Approach
- Cleaned and tokenized review text
- Used NLP models to classify positive/neutral/negative sentiment
- Built product similarity and co-purchase recommender
- Applied hybrid filtering logic for personalization
Results
- 87% accuracy in sentiment classification
- Delivered product recommendations with contextual relevance
- Provided actionable insights for e-commerce UX improvements
👉 View on GitHub