Episode 3 - Stephanie Yang on Foursquare's Rating System

Have you ever wondered how Foursquare rates all of the places in the world? Today I am releasing a discussion with Stephanie Yang about how we used machine learning and sentiment analysis to build Foursquare's critically acclaimed venue rating system.

Here's the Foursquare Engineering Blog post that we recently published. If you are interested in working with me and Stephanie on this and similar projects, particularly if you have a background in software engineering and machine learning, check out the Foursquare jobs page and also email the show with any questions that you have.

Here's a recent article on my blog also about Foursquare ratings, and here's a post about my discovery of different like and dislike rates by language.

Follow Stephanie on Twitter and also check out her blog What Does the Quant Say. That crazy math paper from 2008 that I mentioned can be found here. You need to check out the latex work on pages 5 and 6.

A couple years ago, Stephanie and I collaborated on a poster about trending venues in Foursquare, which is an important component of the ratings for new venues. The paper is here. And the poster itself: Recsys 2016 Poster_ Hot Venues

Towards the end I mentioned some specific venues when it comes to ratings.

In New York, the Museum of Modern Art and the Brooklyn Museum. The best-in-state Central Park. In San Francisco, the Bi-Rite Creamery Ice Cream shop.

On the more negative side, there's New York City Post Offices on 14th street and 146th street. In Moscow, the Passport and Visa Center, and in Bayan Lepas Malasia, the Pizza Hut in Queensbay Mall that can't seem to catch a break!

And from the old scrapbook of different versions of ratings we were playing around with, here is an interesting list of the top sentiment venue in each state, district, and Canadian province. It still needs to be linkified!