A very interesting approach to this is to use self organizing maps (SOM, a kind of a neural net) to organize content based on similarities. The neural net research unit at Helsinki University of Technology has applied such technology to very large volumes of data with good results.
The SOM approach uses the content itself instead of links to sort the data. It groups similar content together and can adapt to almost any kind of data. Best of all, the operation of SOMs is fully automatic!
An example that contains over a million Usenet messages is at:
Slashdot Mirror
A very interesting approach to this is to use self organizing maps (SOM, a kind of a neural net) to organize content based on similarities. The neural net research unit at Helsinki University of Technology has applied such technology to very large volumes of data with good results.
o t.html
The SOM approach uses the content itself instead of links to sort the data. It groups similar content together and can adapt to almost any kind of data. Best of all, the operation of SOMs is fully automatic!
An example that contains over a million Usenet messages is at:
http://websom.hut.fi/websom/milliondemo/html/ro
With a different coloring you could think of it as a collection of islands of knowledge in the sea of information.