Insight Project 2019B
The purpose of my project is to help users of Yelp quickly and easily see what reviewers are talking about for a particular restaurant. In a single review, users could mention multiple aspects of the restaurant, for example talking about the quality of the food, the noise level, the decor, the level of service, etc., obscuring what goes into their ultimate star rating. In this iteration of my project, users can input a Yelp url for a business and immediately see the breakdown of the comments based on the topics talked about, can filter comments for the topics of interest, and can see a rudimentary breakdown of the rating by those same topics. I encourage you to try it out!
The files are organized into folders roughly corresponding to parts of the data science pipeline.
- 0_data (not uploaded) - This folder contains all the raw and intermediary data files.
- 1_data_processing - This folder contains the code to transform the raw data to other usable forms for different tasks.
- 2_data_analysis - This folder contains scripts for functions that often need to be repeated. It also contains jupyter notebooks for ad hoc data exploration and data visualization.
- 3_reports - This folder contains some logs and figures.
- 4_models (not uploaded) - This folder contains the resulting models.
- 5_flask_app - This folder contains the code to run the web app.