This project is part of milestone project of the MADS Master program at University of Michigan.
Goodreads is an online catalog where users can search, review, and register books. This website also allows for quote and annotation searches. With so much text information offered on the website, our team was motivated to perform a sentiment analysis on Goodread reviews.
To do this, we used datasets from the goodreads website UCSD Book Graph website (See requested citations below). Our goal for this project was to see which machine learning method, supervised or unsupervised, was best for predicting sentiment from the review text.
The full details of the analysis is in the report pdf file.
Please note that the project report is being reviewed by the instructor team.