spam with random words
relson at osagesoftware.com
Wed Jan 14 15:17:45 EST 2004
On Wed, 14 Jan 2004 21:06:45 +0100
Piotr KUCHARSKI <chopin at sgh.waw.pl> wrote:
> On Mon, Jan 12, 2004 at 10:12:13AM +0000, T. Horsnell (tsh) wrote:
> > stream of random dictionary words thus:
> Your message was tagged
> X-Bogosity: Yes, tests=bogofilter, spamicity=0.997873,
> It would have landed in the spam folder, if not for an escape hook
> for this mailing list I had.
> So I guess random words do not help the spammers.
That's what we've been saying for a long time. All the different
spellings for viagra are just more clues to make life easier for
bayesian filters like bogofilter.
I wish I could take credit for the idea, but guess I have to settle for
partial credit for a useful implementation.
Try scoring the message using "-vv" and again using "-vvv". The first
will give you a histogram and the second will show exactly what tokens
bogofilter found and their associated scores.
More information about the Bogofilter