BBQ Data Analysis Through Yelp Reviews
The skills I demoed here can be learned through taking Data Science with Machine Learning bootcamp with NYC Data Science Academy.
The data collected is from the ten most populated cities in the US with a focus on BBQ restaurants sorted by number of reviews. Data for the top 100 restaurants was collected for each city. Yelp data was collected using Scrapy. The purpose of this analysis is to determine which city will provide a traveler with the best BBQ experience.
Data on Rating vs Number of Reviews
To start I wanted to compare the number of reviews per restaurant to with the rating for that restaurant. This graph shows that as the review count increases the restaurant becomes more likely to have a good review. With the exception that perfect 5 star reviews are rare. It should also be considered that if a restaurant does have many bad reviews they are likely to not use Yelp at all due to loss of business.
Data on Cost and Ratings
The cost for BBQ seems to be similar throughout most of the cities. New York does have noticeably more 3 ratings for cost.
Adding cost to the figure comparing rating to number of reviews shows that as the cost of a BBQ establishment increases it is less likely to obtain bad reviews. Once the cost rating reaches 3 there becomes no reviews below 3.5 stars. It should still be noted that even the lowest cost restaurants do have the number of 4 to 5 star reviews.
Data Distribution of Overall Ratings
Next I established a good and bad range for reviews. The most common review rating is 4 stars. This leads me to believe that a 4 star review should be considered a good review. Anything more than 4 stars will be considered good and anything less than 3.5 stars is seen as very negative due to the low number of occurrences. This leaves a 3.5 star rating as a slightly negative review, which is still overall positive.
Ratings by Location
Comparing the distribution of ratings for each city shows what locations are more likely to have a well reviewed restaurant. It can be seen that San Jose CA and New York NY have a large number of 3.5 star reviews and more 3 star reviews than most other cities. While San Diego CA, San Antonio TX, and Los Angeles CA have an above average number of 4.5 star reviews with the most 5 star reviews occurring in Los Angeles. The below information leads to the decision that Los Angeles CA will provide the most enjoyable BBQ experience.
Future Work
I hope to revisit this project and add a number of components. Firstly, I would like to spend some time creating a criteria for what cities to compare instead of basing this choice only on population. Secondly, more categories of food would be considered. Finally, I would like to add sentiment analysis of the individual restaurant reviewer text and ratings that were collected. This could offer a look at how kind different locations are relative to each other.
About Author
