A Tale of Two Sisters
Tom Anderson
tanderso at oac-design.com
Mon Jul 25 14:57:26 CEST 2005
Use "bogofilter -vvv" to figure out which tokens are causing the problem.
Tom
----- Original Message -----
From: "JoeHill" <joehill at sympatico.ca>
To: <bogofilter at bogofilter.org>
Sent: Sunday, July 24, 2005 7:55 PM
Subject: A Tale of Two Sisters
>
> Hi all,
>
> As you may gather from the subject this is about my two sisters, whom I've
> been
> trying to train bogofilter as 'ham' (oh, the irony).
>
> It was kinda hit and miss for both, so I sat down today and did training
> on a
> whole *whack* of mail from each sister (about 20MB each)
>
> One seems to be going okay:
>
> joehill at node3:~/mail$ bogofilter -vv < sister1
> X-Bogosity: Spam, tests=bogofilter, spamicity=0.500000, version=0.94.4
> int cnt prob spamicity histogram
> 0.00 3173 0.010475 0.005447
> ################################################
> 0.10 142 0.113748 0.008133 ###
> 0.20 0 0.000000 0.008133
> 0.30 0 0.000000 0.008133
> 0.40 0 0.000000 0.008133
> 0.50 0 0.000000 0.008133
> 0.60 0 0.000000 0.008133
> 0.70 0 0.000000 0.008133
> 0.80 339 0.886811 0.107046 ######
> 0.90 3035 0.978600 0.491909
> ##############################################
> joehill at node3:~/mail$ bogofilter -n < sister1
> joehill at node3:~/mail$ bogofilter -vv < sister1
> X-Bogosity: Ham, tests=bogofilter, spamicity=0.000000, version=0.94.4
> int cnt prob spamicity histogram
> 0.00 12691 0.007783 0.006979
> ################################################
> 0.10 148 0.113356 0.008205 #
> 0.20 0 0.000000 0.008205
> 0.30 0 0.000000 0.008205
> 0.40 0 0.000000 0.008205
> 0.50 0 0.000000 0.008205
> 0.60 0 0.000000 0.008205
> 0.70 0 0.000000 0.008205
> 0.80 339 0.888703 0.055743 ##
> 0.90 1123 0.950223 0.220665 #####
>
> However, sister #2 is not faring as well:
>
>
> joehill at node3:~/mail$ bogofilter -vv < sister2
> X-Bogosity: Spam, tests=bogofilter, spamicity=0.500000, version=0.94.4
> int cnt prob spamicity histogram
> 0.00 1943 0.009352 0.007266
> ################################################
> 0.10 49 0.114453 0.009570 ##
> 0.20 0 0.000000 0.009570
> 0.30 0 0.000000 0.009570
> 0.40 0 0.000000 0.009570
> 0.50 0 0.000000 0.009570
> 0.60 0 0.000000 0.009570
> 0.70 0 0.000000 0.009570
> 0.80 84 0.887220 0.073546 ###
> 0.90 487 0.972303 0.380932 #############
> joehill at node3:~/mail$ bogofilter -n < sister2
> joehill at node3:~/mail$ bogofilter -vv < sister2
> X-Bogosity: Spam, tests=bogofilter, spamicity=0.500000, version=0.94.4
> int cnt prob spamicity histogram
> 0.00 1947 0.006888 0.005492
> ################################################
> 0.10 60 0.114236 0.008374 ##
> 0.20 0 0.000000 0.008374
> 0.30 0 0.000000 0.008374
> 0.40 0 0.000000 0.008374
> 0.50 0 0.000000 0.008374
> 0.60 0 0.000000 0.008374
> 0.70 0 0.000000 0.008374
> 0.80 72 0.887940 0.065235 ##
> 0.90 421 0.978003 0.368686 ###########
>
> Why would one seem to 'learn', and not the other (once again, irony, if
> you knew
> my sisters :-\)
>
> --
> JoeHill / RLU #282046 / www.freeyourmachine.org
> +++++++++++++++++++++++++++
> "Truly, I say to you, as you did it to one of the least of these my
> brethren,
> you did it to me." -- Jesus Christ
> _______________________________________________
> Bogofilter mailing list
> Bogofilter at bogofilter.org
> http://www.bogofilter.org/mailman/listinfo/bogofilter
>
>
>
More information about the Bogofilter
mailing list