Taste graph fürs Web: Hunch

Hunch is like Amazon or Netflix recommendations, but for everything. To build the system, we needed a vast amount of data about people, their tastes and preferences, and the things that they like. Our ambition is to build the „taste graph“ of the web, analogous to the „social graph“, mapping every person on the web to every entity on the web (be that a hotel, or a tennis racket, poet, cookbook, etc.) and their affinity for that entity. You can see how you could take your Hunch taste profile and apply it to just about anything — to find you a hotel in Dallas, or your true love on a dating site, or things you like on eBay or Etsy. We’ve spent the time since launch gathering that data, and we can now really build this out. You’ll be seeing a bunch of new features coming soon!

Weil ihr neues Projekt Hunch die nächste Finanzierungsrunde eingeleitet hat, erklärt Flickr-Erfinderin Caterina Fake was so toll ist an dem Geschmacksratgeber, der die Einschätzung der Vielen zu einem Orientierungsmaßstab im unübersichtlichen Web erheben will – sozusagen als soziales Wissen für Geschmacksfragen: