Julian Cole

Head of Comms Planning at BBDO NY



UPDATE 3 - Bourdain's Yelp Challenge

'I've long believed that good food, good eating, is all about risk.' Anthony Bourdain

There is no bigger risk then going againist popular opinion and going head strong into the 1 star hell holes of Yelp reviewed restaurants in your city. 1 star reviews on Yelp also just happen to be some of the best reads going around.  

This site would look to highlight the restaurants and their passionate Yelp Reviewers of the highest 1 starred Yelp reviewed restaurants in your city.

So you would type in the zipcode of your city and it would tell you what restaurants everyone hates. 

Some of the algorithms I think I would build in is; that there has to be over 50 reviews of the restaurant and that it is on the restaurant with the highest ratio of 1 star reviews againist 5 star reviews.


I have run into a little bit of a brick wall. I can't actually search by lowest rated comments using the Yelp API. 



Yelp API is out of order but I think I have found a bigger and more robust API in Goolge Places. They also have reviews and people could pretty animated so I think the idea is still good. Might need to take Yelp out of the title. 


Also found this piece on pulling Google Places into Rails3 - YEW!!!! 




Thanks to David for answering my question about whether it would be cool going ahead with the name Bourdain. I am cool to change it if I get a cease and desist letter. 

In terms of progress, I was pretty excited about using Google Places API for this project, it looked like a great API as it is was well documented however what I found was that there was not a lot of flexibility when it came to ordering your search, it was either in distance or a mix of distance and ratings (I couldn't even get close to lowest ratings. It also came in maps, my coding skills are definitely on the amateur side so I went back to the drawing board. 

I was having a think back on a previous project I had worked on twantrum.com where we used search terms as the guide for how bad someone was cracking the shits at a brand. I thought this could actually work for this project, instead of looking at the ratings look for the bad words people are using to describe the meal. So I typed in food poisoning and got some great results back from Yelp.


So I am back on Yelp, I even did a quick test and was able to pull results back for looking at 'food poisoning' in Williamsburg, Brooklyn. http://coleand.co/geo2/

I have also created some wireframes, I got the idea to use inVision from Illya's project Decisions. Let me know what you think of the wireframes and if you have any builds.  https://projects.invisionapp.com/share/N79I9OPC#/4085594

My next job will be to get the results to look similar to how they are presented on the Yelp Search Results instead of just the code. I also will have to look at getting it over into Rails soon, I feel like I am leaning on my small JavaScript knowledge at the moment to keep this project going. 


Please sign in or sign up to comment.