Patrick Butts wrote: > Fritz Borgstedt wrote: >>> I'm at work now and can't look at previous emails, but wasn't OP's >>> original problem that he had a really low weight because he had way >>> more >>> ham than spam? I don't understand why these settings wouldn't solve >>> the >>> problem. Enlighten me, please, Fritz because I'm obviously confused. >>> >>> >> If you use numbers as filenames (what you should), the maximum number >> of files in spam folder and notspam folder is limited by >maxfiles< >> eg. 15000. >> To change the frequency does not change this maximum number. It does >> change the <frequency> of collecting. So after your collection has >> reached 15000, there is no change in number of files for spam and >> notspam. >> >> fritz >> > Wouldn't this be better then for a installation that isn't mature? For > instance, I only have about 2000 spam in the directory and 500 nonspam > so it seems like this will help me keep the weight down until my > installation matures (reaches total files). I don't want to sound > argumentative. I'm just trying to reach an understanding here.
With numbers that low I would just turn off Bayesian and wait for the corpus to build up. I don't think that you have to reach the maxFiles limit to have a mature corpus, you could probably get away with around 5000 in each folder as long as you were using the reporting to properly report false positives/negatives. The max files setting is so that the size does not grow too large and also so that the corpus can be kept fresh by automatically removing old contributions. How long has your ASSP been running? Kevin ------------------------------------------------------------------------- This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ _______________________________________________ Assp-user mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/assp-user
