From: mailbox@dmitry-kazakov.de   
      
   On Sat, 19 Jun 2004 22:41:03 +0900, Doan Son wrote:   
      
   > I am finding some related papers to represent a tex by a fuzzy set or   
   > based on fuzzy theory ? Can it be represent instead of statistical   
   > approach up to now ? The application maybe in text categorization, text   
   > retrieval, etc.   
      
   I do not think that text should be represented by a fuzzy set. There is   
   IMO little sense in that. I presume that the text is exact (not a   
   result of some recognition process). Also the text is known, though   
   usually it is to big for any direct analysis. Moreover it can be little   
   done with texts except for simple editing opreations (here I assume that   
   we do not talk about translation, compilation and other stuff like that).   
   This is why instead of text one quite often uses its features.   
   Which could be word counts, layout, used fonts, spelling errors etc. These   
   features can be crisp, statistical and of course fuzzy. And fuzzy can be   
   very useful here. For example, people recognize spam without counting   
   words. The way they do it is in my view rather fuzzy than statistical.   
      
   --   
   Regards,   
   Dmitry A. Kazakov   
   http://www.dmitry-kazakov.de   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)