[Spambayes] Confusion about Unix or Linus documentation

akonstam at trinity.edu akonstam at trinity.edu
Wed Jul 21 23:51:12 CEST 2004


On Wed, Jul 21, 2004 at 12:44:38PM -0500, Skip Montanaro wrote:
> 
>     Aaron> In the .procmailrc file the spam is put in a file
>     Aaron> $HOME/Mail/spam.  In the line that illustrates the training of
>     Aaron> spambayes using: sb_mboxtrain.py the file used to train spambayes
>     Aaron> is also $HOME/Mail/spam. It is confuses me that these files are
>     Aaron> the same.  Is it really usefull to retrain spambayes on the files
>     Aaron> it has identified as spam or should these files have different
>     Aaron> names as I have done with spamassassin?
> 
> sb_mboxtrain.py tags messages it's already trained on so when it revisits
> $HOME/Mail/spam at a later time it will only train on messages that have
> been added since the last run.  You can run it with the -f flag to force it
> to retrain on all messages.  Note however that you need to lock access to
> your file since both procmail and sb_mboxtrain.py will want to write to it.
> 
> Skip
I understand that sb_mboxtrain.py will recognize the
messages it has been already been used in training 
but that is not what I am asking. If the messages that the user has
identified as spam are placed in the same file as the messages that
spambayes has already identified as spam, then we would be training
on messages that have never been used in training but spambayes has
already identified as spam.

Is this a useful thing to so?
-- 
-------------------------------------------
Aaron Konstam
Computer Science
Trinity University
One Trinity Place.
San Antonio, TX 78212-7200

telephone: (210)-999-7484
email:akonstam at trinity.edu


More information about the Spambayes mailing list