Jesse> I'm pretty sure I've seen discussion on the list of a training
Jesse> regimen that involved discarding aged messages...
I use a so-called "train to exhaustion" regime (*). One part of my current
implementation is to always process messages from newest to oldest. That
way older messages carry less weight (they will more likely be scored
correctly because newer messages have already contributed their contents to
the training database). Is that what you were thinking of?
Skip
(*) http://www.garyrobinson.net/2004/02/spam_filtering_.html
http://cvs.sourceforge.net/viewcvs.py/spambayes/spambayes/contrib/tte.py?rev=1.16&view=markup
_______________________________________________
[email protected]
http://mail.python.org/mailman/listinfo/spambayes
Check the FAQ before asking: http://spambayes.sf.net/faq.html
