The Speculist: The Wisdom of the Crowd Builds Amazing Maps


Live to see it.

« Wolfram/Alpha Webcast | Main | Flu Season, sort of »

The Wisdom of the Crowd Builds Amazing Maps

If this story isn't a premo example of the power of the wisdom of the crowd, or self-organizing processes, I don't know what could be:

Billions of photos have now been uploaded to the internet, and many are tagged with text descriptions. Some are even geotagged – stamped with the latitude and longitude coordinates at which the image was taken. David Crandall and colleagues at Cornell University in Ithaca, New York, analysed the data attached to 35 million photographs uploaded to the Flickr website to create accurate global and city maps and identify popular snapping sites.

The enormous dataset provides a global picture of "what the world is paying attention to", the researchers say. They ran statistical analyses to identify the more important clusters on each map. Next they analysed the text tags added to photographs in those clusters, as well as key visual features from each image, to automatically find the world's most interesting tourist sites.

According to Flickr, New York is the world's most photographed city. But London contains four of the seven most photographed landmarks in the world – Trafalgar Square, the Tate Modern art gallery, Big Ben and the London Eye. Some bizarre results emerged – the Apple Store in Manhattan is the fifth-most photographed place in the city.

Now just imagine what Wolfram/Alpha could do with this data. Wow!


It would be interesting to see this data graphed in some fashion so as to display changing trends in interest over time. A further interesting experiment would be to publicly solicit Flicker users to publish "themed" images from disparate locales. The variety of unique takes on a given topic would be fascinating, I think.

I gotta say; a source like the Wolfram/Alpha you wrote on earlier would absolutely pollute any such effort I expect.
Since Alpha is intended to be "the" definitive source, any deviation from an Alpha result would have to be "wrong" or at best sub-optimal, don't you think? Not saying the project won't result in an improved data search mechanism, but it absolutely will result in a much more constrained experience. How many times in any of our lives have their been multiple "right" answers depending on highly individual variables? It will be instructive to discover how Mr. Wolfram has dealt with this.

I noticed in the map of the most photographed places on the European continent, the densest areas closely resemble the Kingdom of Lotharingia. It's amazing how ninth-century politics is still relevant.

Post a comment

(Comments are moderated, and sometimes they take a while to appear. Thanks for waiting.)

Be a Speculist

Share your thoughts on the future with more than


Speculist readers. Write to us at:

(More details here.)



Powered by
Movable Type 3.2