I need to recommend method calls that are commonly used with a given   
   method call in Java Code. I do not have any user specific data. All I   
   have is a set of methods that get used together by browsing   
   repositories.   
   ex: A, B,C,D,E,F are method calls in Java   
   pass 1: A, B,C are used together   
   pass 2: A, B   
   pass 3: D, E,F   
   pass 4:...   
   I would like to recommend methos B and C to a user who is using method   
   A.   
   This is a very small portion of my project and I was hoping to find   
   some simple technique to implement this. I thought of using kmeans but   
   am unsure as to how I would set up the distance metric. All i have is   
   the number of time method X occurs with method Y (assuming I am storing   
   this in a DB somewhere). Any ideas as to how I could go about doing   
   this. I am okay with using a fixed predetermined number of clusters,   
   incase k means or knn is doable. Or is storing the methods i db and   
   running some sql queries the way to go?   
   Appreciate the help.   
   Thanks!   
   A   
      
   [ 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)