There’s been much talk recently about how flickr could improve the display and number of photos(dots) on the map. Most of which boils down to server side clustering. But what’s the best way to make clusters? How should we group photos, and which photo best represents that cluster. Luckily for us those same smart folks over at Yahoo! Research Berkeley have been giving this much thought. For a quick overview I highly recommend this slide show.
Part of this clustering can be seen in the above lovely tag map, created by analyzing the locations and tags of photos from flickr. I particularly like the way you can see where the BBC is based. For more info on this see the post; Tag Maps: Visualizing the Crowd’s “Mental Map†Using Flickr Geotagged Images from the Yahoo! Research Berkeley Blog. There’s far more hardcore juicy details in the official paper thing (pdf 644kb), you know, those things academics write.
You can be pretty sure that some of this will be pulled into flickr one way or another.
I’ve experimented a bit with this before. The demo below plots tags anywhere the map is browsed. It actually doesn’t include the underlying maps .. and in some places it’s possible to navigate purely by flickr tags! Other places there’s absurdities and serendipities.
http://brainoff.com/openplan/flickr/
A big improvement would be to use clustering for better visibility.