Yelp_Reviews_Analysis

This project have two goals.
Goal1: Provide data-driven, actionable decisions to owners in order to improve their ratings in Yelp.
Goal2: Provide a prediction model to predict the ratings of reviews based on the text and related attributes and submit the predictions to Kaggle.

Code folder:

Goal1: Regular Expression, NLTK(pos_tag, word_tokenize), Logistic Regreesion, tf-idf, Linear Regression
Goal2: LSTM(keras)

Slide folder:

The whole idea of this project.

Result folder:

The result of goals respectively.