Re: [Evolution] My two problemos

[guenther <guenther rudersport de>]

> > Problemo #2:
> > I use spamassassin, and would like to find a way to pipe an email
> > message to 'sa-learn' to keep my spam filter trained. Is this feature
> > available?
>
> I don't know about this, because normally you "sa-learn" spams that
> Spamassassin does not recongnize as spam, and so it is going right into
> a "normal" inbox (not the SPAM Inbox).

It's generally a good idea, to learn every HAM and SPAM, regardless if
SA recognized it or not.

The point is, recognizing a mail as SPAM does not necessarily mean, this
particular mail is auto-learned. You constantly have to feed new mails
to the Bayes database to improve it.

Even more, you have to do the same for HAMs as well, as you may end with
a badly biased Bayes database otherwise. Bayes even won't kick in,
before there are 200 SPAMs and 200 HAMs (sic) learned.


> I am moving manually this spam in a temporary folder, named ZPAM, as
> spamassassin is catching 99% of my spam, so this is not a hard job.
> Then, once a while, I am running this script that I found on this
> mailing list.
>
> #!/bin/sh
> echo Learning about spam
> # On a single line below
> sa-learn --spam --showdots --mbox /home/your_user/evolution/local/ZPAM/mbox
> echo Rebuilding database
> sa-learn --rebuild
> echo Training complete!

Add a --no-rebuild option to the first sa-learn command, or you safely
can drop the --rebuild command AFAIK.

...guenther


-- 
char *t="\10pse\0r\0dtu\0  ghno\x4e\xc8\x79\xf4\xab\x51\x8a\x10\xf4\xf4\xc4";
main(){ char h,m=h=*t++,*x=t+2*h,c,i,l=*x,s=0; for (i=0;i<l;i++){ i%8? c<<=1:
(c=*++x); c&128 && (s+=h); if (!(h>>=1)||!t[s+h]){ putchar(t[s]);h=m;s=0; }}}