Group Members

Fabio Fleitas, Guillermo Gutierrez, Jake Hart, Ian Sibner


Yelp reviews, even as helpful as consumers find them, have several inherent aws. Reviewers may be biased, inconsistent, or simply value different qualities than other Yelp users. In addition, Yelp assigns star ratings to businesses based upon a simple average of the reviews of the business. In addition to the aforementioned concerns about accuracy of reviews, the simple average method allows the business ratings to be subject to small sample variation and sensitive to in uential outliers. Finally, the negative skew of Yelp data makes it difficult to evaluate which businesses are truly elite. Thus, our group set out to create a more accurate way for consumers to judge businesses based upon Yelp data. Using a robust technical infrastructure, Bayesian statistical methodology, and the provided Yelp data, we set out to achieve this goal.