YelpReplicate

Replication code for WWW'19 Paper ``Is Yelp Actually Cleaning Up the Restaurant Industry? A Re-Analysis on the Relative Usefulness of Consumer Reviews''

Documentation

This repository contains all the corresponding code to recreate tables and figures corresponding to the re-analysis portion in ``Is Yelp Actually Cleaning Up the Restaurant Industry? A Re-Analysis on the Relative Usefulness of Consumer Reviews''.

Directions

1. Save the raw Yelp data called 'instances_mergerd_seattle.csv' from Kang et al. 2013 [link].

2. How to run code where folder organization is briefly described below:

• ./data/: includes a link to the original data subsetted to routine-inspections only as used for the re-analysis portion of the paper and includes folders for storing features from the routine-only subsetted dataset.
• ./code/0_original_analysis_restaurant_hygiene/: placeholder for including original Kang et al. code.
• ./code/1_re-analysis/: code for re-analysis.
• ./figs/: folder for all figures
• ./appendix/: folder contains document with supplemental figures that would not fit in main paper.

All code was writted and tested for Python 2.7.12.

Authors

• Kristen M. Altenburger, kaltenb@stanford.edu
• Daniel E. Ho, dho@law.stanford.edu