From: marina@sapir.us   
      
   It is not correct to say that the proposed method requires an   
   assumption that all edits are equally important. If you do assume that   
   all edit are equally important, all you need is voting: just count how   
   many edits (transformations) of all kinds there are in each string.   
      
   This method works precisely when one can not make any justified   
   assumptions about importance of edits. This was the question Scott and   
   Jim asked.   
      
   Now, you change the question. You are saying: what, if one  has reasons   
   to believe, that particular edits (patterns, transformations) are more   
   important than other.   
      
   I find it fascinating, that Wittkowski, actually, considers such a task   
   in his paper. He has an example, when one needs to compare performance   
   of different countries on the Olimpiade. To do it, he takes into   
   account that gold medals are more important than silver ones. Please,   
   go directly to his paper, if you are interested.   
      
   The distribution I am after is generalization of the concept of   
   U-statistics. Wittkowski talks about it too. But not much.   My guess   
   is I need to help myself here.   
      
   Marina Sapir   
   http://sapir.us   
      
   [ comp.ai is moderated.  To submit, just post and be patient, or if ]   
   [ that fails mail your article to , and ]   
   [ ask your news administrator to fix the problems with your system. ]   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)