Shapes from the crowd

Here's an excellent post from Christian about flickr (geo-) shapes - derived from crowd-sourced tags.

Shapes from the crowd - Christian Spanring

The 2 examples he gives (Vienna boundaries, and Vienna inner city) to me are typical examples of the "wisdom of crowds" effect: quite accurate with a large enough sample and good mixture of people within the crowd (Vienna boundaries) and sometimes "totally off" with a specific narrow topic and (therefore) a badly composed crowd (Inner city vs Stephansdom).

It also highlights the problem with tags, that - especially on flickr - it is unclear what the are used for: the place where the picture was taken from, or the motif...

Thanks.