The repository contains two parts of data analysis based on the Yelp dataset. The first part is building a recommedation system for the merchants and users on Yelp based on the business attributes and the reviews. The second part is building a deep learning model to predict the ratings of each review based on the five million training text reviews and their related ratings.
-
a folder called code, containing two folders:a folder called recommendation system, which contains all of our work in part one, a folder called reviews predict ratings, which contains all of our work in part two.
-
a folder called data, contains raw data in json and cleaned data in csv.
-
a folder called report, which contains our report combining our results both in part one and part two.
-
a folder called img, which contains the images to load in our report.
-
Some snippets of our app.
