[Spambayes] Good messages vs. spam

[Glenn Merritts]

I trained the system using the folders with good messages and folder with
bad messages as recommended. When I finished, I had a good balance of good
vs. spam per the manager. As subsequent messages were received and went into
the suspect folder, I sorted the file by spam % and physically deleted the
some of the lower valued messages as spam and then had the filter re
evaluate the remaining messages; it would pick up additional spam and leave
others. I continued this process until the suspect folder was empty. I have
continued this procedure since my initial installation, but now the manager
says my good vs. bad is out of balance. I have lowered the spam limits and
have not been getting any good email in the junk or suspect folder so I am
pretty confident I have things configured correctly; except for the out of
balance warning.

 

Since the majority of my email is spam, or suspect and using the ranking
procedure above, I still have to physically designate a certain amount of
email as spam every time I clear the folder. I rarely have to designate an
email as good from either the junk or suspect folder. But I am sure my out
of balance condition is caused by having to physically designate email as
spam over time (if I have to designate one email as good, I will have
probably had to designate 20 as bad); the spam count has to grow
disproportional to the good over time. 

 

How to I fix this without starting over? I would use the same procedures to
train and would eventually wind up with the same situation - unless I
completely misunderstood the documentation on how to train the system.

 

Thanks for the help.

 

I am using the November version of the system; my operating system is WinXP
Professional; my email is Outlook (contained in Office XP).

 

 

Glenn P. Merritts

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.python.org/pipermail/spambayes/attachments/20041215/5d604b5c/attachment.html