You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@spamassassin.apache.org by Markus Braun <ch...@hotmail.com> on 2006/01/28 17:00:54 UTC

Spamassassin Spam Header

Hello again,

sorry but iI have still the problem , that i dont understand the spam bayes 
function.
I have SA-learn about 50 emails they are spam. But always the same header as 
spam comes as ham into my mailbox.

e.g.:

The Ultimate Online Pharmaceutical
Software
Software At Low Pr1ce

How can i make that such emails with these headers are marked as spam?

Thanks
marcus

_________________________________________________________________
Sie suchen E-Mails, Dokumente oder Fotos? Die neue MSN Suche Toolbar mit 
Windows-Desktopsuche liefert in sekundenschnelle Ergebnisse. Jetzt neu! 
http://desktop.msn.de/ Jetzt gratis downloaden!


Re: Spamassassin Spam Header

Posted by Markus Braun <ch...@hotmail.com>.
>Are you sure your sa-learning was done as the correct user?
>What user did you use? (root? nobody? The message recipient?)
I logged in as root.
I type sa-learn -spam and then the name of the email file.

>Do you use spamd?
yes spamd is loaded.

>Do you normally pass a -u parameter to either spamd or spamc?
I dont know  :-(, i installed it from apt-get install spamassassin

>What BAYES rules do the messages hit when they are scanned at delivery 
>time?
Where can i look ?


>You can drop these into a .cf file in your site-config directory
>(/etc/mail/spamassassin/ for most people):

Yes i made a file called rule.cf in this directory.

I think i must restart the daemon?

bye
marcus

_________________________________________________________________
Die neue MSN Suche Toolbar mit Windows-Desktopsuche. Suchen Sie gleichzeitig 
im Web, Ihren E-Mails und auf Ihrem PC! Jetzt neu! http://desktop.msn.de/ 
Jetzt gratis downloaden!


Re: Spamassassin Spam Header

Posted by Matt Kettler <mk...@evi-inc.com>.
Markus Braun wrote:
> Hello again,
> 
> sorry but iI have still the problem , that i dont understand the spam
> bayes function.
> I have SA-learn about 50 emails they are spam. But always the same
> header as spam comes as ham into my mailbox.

Are you sure your sa-learning was done as the correct user?

What user did you use? (root? nobody? The message recipient?)
Do you use spamd?
Do you normally pass a -u parameter to either spamd or spamc?

What BAYES rules do the messages hit when they are scanned at delivery time?



> 
> e.g.:
> 
> The Ultimate Online Pharmaceutical
> Software
> Software At Low Pr1ce
> 
> How can i make that such emails with these headers are marked as spam?

Well at least two of those are best dealt with by custom rules.. If you've got a
very specific, often repeated pattern, static rules help a lot and are a
sure-fire way to cover it.

You can drop these into a .cf file in your site-config directory
(/etc/mail/spamassassin/ for most people):

header L_S_ULT_PHARM    Subject =~/\bThe Ultimate Online Pharmaceutical\b/i
score L_S_ULT_PHARM     3.0
describe L_S_ULT_PHARM  spam from the ultimate online pharmacy

header L_S_SW_LOWPR    Subject =~/\bS[o0]ftw[a4]r[e3] At L[o0]w Pr[i1]c[e3]s?\b/i
score L_S_SW_LOWPR     3.0
describe L_S_SW_LOWPR  offers software at low price



Note: careful of the word-wrapping, each of those rules is 3 lines long, for a
total of 6 lines.