Model comparison for rating prediction in Yelp reviews

This project is a comparison betweent sentiment analysis based models and latent factor models for rating prediction using the public Yelp dataset Yelp reviews. We compared the performance of a few models in predicting the rating using other information/features given, such as review text, business location, user id, etc. The performance was measured by the RMSE between the predicted ratings and the actual ratings in reviews. The results were compiled as a short paper in the pdf file.