This github repo provides the code and report for an Amazon rating predictor. The rating predictor is a linear regression model where term frequency - inverse document frequency was used to featurize the review text. The predictor was written to compete in a Kaggle competition which can be found here: https://www.kaggle.com/c/csc2515-rating-prediction/overview
The final model had a mean squared error of 0.57. The report discusses my previous attempts and design decisions for the project.