On 08 Oct 2003, David Krider wrote:
> On Wed, 2003-10-08 at 03:44, Kjetil Kjernsmo wrote:
> > So, I would strongly recommend you just manually save the spam to a
> > folder for some time now, and build your spam database from there.
> Don't forget that spamassassin is building it's heuristics database with
> every email that you get, so it's learning about your ham with every
> good message that you get. Since you're subscribed to a high-volume
> email list, there's a large corpus right there. So, saving off your
> uncaught spam to do a bulk ``sa-learn'' with isn't as hopeless a job as
> it would first appear. I have found that running it against as few as 50
> uncaught spams has dramatically reduced the amount of spam I get in my
> inbox. I had been getting a couple of particular spams per day, and
> after about a week, I taught spamassassin, and now I'm back down to one
> or two spams a week. The bottom line here is not to worry if you don't
> have thousands of spams to feed to spamassassin, you can get it to work
> much better on just tens of them.
I'd say it is at least worth thinking about changing to spamprobe. This
doesn't seem to get as much attention as some of its rivals but I've
found it to be better than the others I've tried. It's faster than
spamassassin and for me at least gets few false negatives and no false
positives at all.
email@example.com || http://www.acampbell.org.uk
using Linux GNU/Debian || for book reviews, electronic
Windows-free zone || books and skeptical articles