Life in the startup lane

I am loving the new features in the new Foursquare 3.0 release. Being the big believer in the wisdom of the crowds that I am, I often rely on the Foursquare tips to help me navigate menus to find the best entrees each time I visit a restaurant. That is why I have been looking forward to using the new social recommendation feature in Fourquare 3.0 which is based is based on tips, to-dos and volume of check-ins. The recommendation feature is described in this excerpt from this Foursquare blog post:

For many, foursquare has been a great way to find out about the places your friends frequent (through check-ins) and learn about specific experiences to seek out (through tips and to-dos). For years we’ve wanted to build a recommendation engine for the real world by turning all the check-ins and tips we’ve seen from you, your friends, and the larger foursquare community into personalized recommendations.

This morning I was playing around with the social recommendation feature and observed that the recommendations are based on the popularity of a place. It looks like “popularity” is determined by the number of check-ins and tips. Unfortunately, Foursquare is not parsing the tips to determine content and context. Check out the following screen-shots for a Walgreens Pharmacy that Foursquare recommended:

Here is a Walgreens recommendation that came up as part of my search results. The tip below it caught my eye so I checked out the rest of the tips. See below.

Both tips for this Walgreens are negative.

For a recommendations, I would suggest that Foursquare parse the tips for sentiment. As a Foursquare recommendation service user, I would like the places that are recommended to me to be not only places that have a high volume of check-ins and tips but also a high volument of positive sentiment tips. If the algorithm is enhancement to include some sentiment analysis, then this Walgreens store would not have been recommended to me.