Hello Mica Mijatovic & everyone else, on 19-Feb-2005 at 23:28 you (Mica Mijatovic) wrote:
>>> Filter against what you *accept*/receive, not against what you do >>> *not*. (-: >> Isn't that exactly what you do when you train a(ny) Bayes filter? > I'll try to say this in other way... When you make a party, you do not > have to check entire city to know who is *not* invited, in order to > decide who of them will be passed in in your beautiful house; you pass > just those with invitations of yours. (-: The fine people who address you > properly. I make a party all the time with my emails :) and all kind of people (mails) come by. I classify the mails and tell them "you may get in" and "you may not get in". I don't worry about the whole city (world email traffic) either, only about those people who get by and want to join the party (mails that are somehow addressed directly to me). Whats so charming is that my colleague the Bayes filter learns which mails I accept, and which not, so I don't have to worry myself. -- Best regards, Alexander (http://www.neurowerx.de - ICQ 238153981) Each waking day is a stage dominated for good or ill, comedy, farce, or tragedy, by a dramatis personae, the `self, and so it will be until the curtain drops. -- Sir Charles Sherrington ________________________________________________ Current version is 3.0.1.33 | 'Using TBUDL' information: http://www.silverstones.com/thebat/TBUDLInfo.html
