Re: sa-learn On Ham On POP server

[Sean Kennedy <sk...@tpno.org>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Wednesday 09 June 2004 14:25, mazur wrote:
> I use SpamAssassin and everything works pretty well.  I am also asking
>  the Bayesian Classifier to learn using sa-learn, and pointing it to
> my resulting spam file.
>
> But, this machine runs a pop server and all users pull their email off
>  regularly.  How can I ask sa-learn to learn from the "ham" also, as
> the docs mention that the bayesian tool will become effective only
> after analyzing 200 spam *and* 200 ham messages.  There isn't a ham
> file to analyze if users are draining the mail files?!
>
> I could see if I were running a IMAP server, but like many, I am not.
>  So what do you do?
>
> Thanks,
> Rob

Archive email as it comes in, sa-learn it at the end of the day, then delete 
it.  

I would let your users know if you plan on doing this, as there are privacy 
concerns they may raise with you.

How do you know what is spam and what is ham, btw?  Do you read all the spam, 
and all the ham, to be sure?
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)

iD4DBQFAx4UjIjyA6vwcN38RAgQ9AJUWGrkxWYkGR4QvLuZhuOYginkaAJsEPgX6
Cj1gt3exL1J4qmECAN67rw==
=wjl+
-----END PGP SIGNATURE-----