[FoRK] Text trees: semi-awesome

[J. Andrew Rogers]

On Feb 14, 2008, at 8:21 AM, Jeff Bone wrote:
> Going to have to dig in and figure out how the clustering occurs,  
> but it seems to me that this kind of destructuring topological sort  
> might be a very, very useful tool in advancing the state-of-the-art  
> in natural language understanding by machines.  (Anybody know if  
> this kind of thing's already being used for that sort of purpose?


It is suffix tree clustering, and has been around in various forms  
since the 1970s.  From the standpoint of natural language processing,  
it is essentially equivalent to text compression methods (IIRC  
something like suffix trees are used in the Burrows-Wheeler transform)  
with all the capabilities and limitations implied.

Cheers,

J. Andrew Rogers