top of page

TEAM

Predicting Product Ratings from Reviews

Anish Joseph, Vishal Bhatoy, Rachel Lopez, Obada Nairat, Eric Malitz

clear.png

Problem: E-commerce companies need accurate predictions of customer satisfaction to improve product recommendations and marketing. However, existing rating data may not fully capture customer sentiment expressed in reviews.

Dataset: Amazon product reviews dataset containing ratings (1-5 stars), price, category, and free-text reviews for various products. (search amazon ratings on kaggle for 2023 Indian Amazon data)
Methods:
1) Develop baseline classification/regression models using only structured data (ratings, price, category)
2) Pre-process and vectorize review text
3) Incorporate review text features into models to determine if it improves predictions
4) Analyze model performance and insights gained from both structured and unstructured data

Potential Impact: More accurate predictions of customer satisfaction can help companies better target marketing and improve customer experience. The project also applies machine learning techniques learned in the course.

Screen Shot 2022-06-03 at 11.31.35 AM.png
github URL
bottom of page