> From: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] On > Behalf Of Tony Meyer > Sent: Saturday, May 13, 2006 11:09 PM > To: Beverly Wertman > Cc: [email protected] > Subject: Re: [Spambayes] Undo Delete as Spam Sender > > > > I need to undo selecting a sender into "delete as spam" > > I had them selected and hit the wrong button. > > Simply open up your spam folder, find the erroneously trained > message, select it, and click the "Recover from Spam" button. > SpamBayes will undo the bad training (and also train as ham), > and move the message back to where it came from. > > =Tony.Meyer
I keep learning things I didn't know about Spambayes training. It appears that you're saying that recovering the message to the inbox does more than just reverse the effect of misclassifying it as spam. I.e., if it "also train[s] as ham", doesn't this mean that the (sole) end result is to add a message to the ham database that wouldn't have been added otherwise? This suggests that misclassifying ham messages and then reversing the misclassification can be used to help redress the balance when the ham/spam ratio gets too low. Or am I reading more into this than you intended? Bob Information Technology Department, City of Cambridge 831 Massachusetts Ave, Cambridge MA 02139 * 617-349-4217 * fax 617-349-6165 _______________________________________________ [email protected] http://mail.python.org/mailman/listinfo/spambayes Check the FAQ before asking: http://spambayes.sf.net/faq.html
