Understanding tuning results
Boris 'pi' Piwinger
3.14 at logic.univie.ac.at
Fri Jun 6 11:14:12 CEST 2003
David Relson wrote:
>> >That looks much nicer ! I'd choose 1e-2, 0.475, 0.503000 from the equally
>> >good results and see how it goes over the next few days. Since I've used
>> >1e-2 for robs for a while, it'd be comfortable to continue using that value.
>>
>>I just changed it and got my first spam:-(
>
> Why the sad face? I enjoy getting up in the morning and seeing the 50 to
> 100 spam that came in overnight - all sitting in their neat little folder
> awaiting their trip to the bit bucket.
I meant the first spam get through. it now turns out that I
missed to paste the spam_cutoff=0.503 line. Anyhow, I got
four spam mails come through starting with the one metioned
above:
$ bogofilter -vM<spam
X-Bogosity: Ham, spamicity=0.476, version=0.13.6.2/fisher
X-Bogosity: Ham, spamicity=0.501, version=0.13.6.2/fisher
X-Bogosity: Spam, spamicity=0.721, version=0.13.6.2/fisher
X-Bogosity: Spam, spamicity=0.668, version=0.13.6.2/fisher
$ bogofilter -c ~/.bogofilter.old.cf -vM<spam
X-Bogosity: Ham, spamicity=0.007, version=0.13.6.2/fisher
X-Bogosity: Spam, spamicity=0.944, version=0.13.6.2/fisher
X-Bogosity: Spam, spamicity=1.000, version=0.13.6.2/fisher
X-Bogosity: Ham, spamicity=0.510, version=0.13.6.2/fisher
So both my old config and the new one suggested by the
tuning script do fail to catch two of them.
pi
More information about the Bogofilter
mailing list