You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@spamassassin.apache.org by Marcin Mirosław <ma...@mejor.pl> on 2012/04/20 23:20:11 UTC
use_bayes=0 completly disables report function
Hello,
i've notice when i set use_bayes 0 then spamc -C report stops to work.
I've got in log: spamd: Can't call method "learn" on an undefined value
at /usr/lib64/perl5/vendor_perl/5.12.4/Mail/SpamAs
sassin/PerMsgLearner.pm line 111, <GEN187> line 117.
Whay i've to use bayes when i want to report spam to
spamcop/dcc/pyzor/razor/etc ? Couldn't it be independent?
Regards,
Marcin
Re: use_bayes=0 completly disables report function
Posted by Marcin Mirosław <ma...@mejor.pl>.
W dniu 2012-04-21 07:38, Marcin Mirosław pisze:
> I'm guessing, as for now, it's the best to set
> connection to non existent database (disadvantage is high number of
> error messages in log).
I tried, it looks none rror messages appears in logs. Fine for me:)
Regards,
Marcin
Bayes for large heterogeneous user-base (was Re: use_bayes=0
completly disables report function)
Posted by "David F. Skoll" <df...@roaringpenguin.com>.
On Sat, 21 Apr 2012 14:40:42 -0500 (CDT)
Dave Funk <db...@engineering.uiowa.edu> wrote:
> > Beyes does not make much sense in a multi-user, diverse community
> > such as my university department. Makes sense here (small company;
> > small user base)
> I'll have to disagree with that, as a person running a mail server
> for a university college (organizational unit bigger than a
> department) which has thousands of users. Bayes may not be as deadly
> accurate as it would in a totally homogeneous environment but still
> worthwhile.
+1. Our (commercial) solution includes a nightly-updated Bayes corpus
that includes tokens from a couple of million messages from more than
a million end users and it's still extremely accurate at picking out
spam. As you wrote, ham is all over the place, but spam tends to look
the same.
Regards,
David.
Re: use_bayes=0 completly disables report function
Posted by Dave Funk <db...@engineering.uiowa.edu>.
On Sat, 21 Apr 2012, John ffitch wrote:
> On Sat, 21 Apr 2012, RW wrote:
>
>>
>> That's what use_bayes 0 is for.
>>
>> Unless you really have to, turning-off Bayes is a bad idea. The FP rate
>> is much higher without it.
>>
>
> Beyes does not make much sense in a multi-user, diverse community such as my
> university department. Makes sense here (small company; small user base)
>
> ==John ff
I'll have to disagree with that, as a person running a mail server for a
university college (organizational unit bigger than a department) which has
thousands of users. Bayes may not be as deadly accurate as it would in a
totally homogeneous environment but still worthwhile.
Altho the ham may be all over the ball-park, the spam tends to cluster in
waves. Also Bayes is as much about the headers as it is the body and so
will find commonalities that aren't obvious at first glance.
FWIW I run a site-wide bayes with auto-learning and occasional hand
training. Maybe not optimal but still worth doing and doesn't need much
attention. Over the past 9 years I've had to discard my Bayes database and
start from fresh (due to going totally off the rails) -once-.
--
Dave Funk University of Iowa
<dbfunk (at) engineering.uiowa.edu> College of Engineering
319/335-5751 FAX: 319/384-0549 1256 Seamans Center
Sys_admin/Postmaster/cell_admin Iowa City, IA 52242-1527
#include <std_disclaimer.h>
Better is not better, 'standard' is better. B{
Re: use_bayes=0 completly disables report function
Posted by Benny Pedersen <me...@junc.org>.
Den 2012-04-21 15:23, John ffitch skrev:
> On Sat, 21 Apr 2012, RW wrote:
>> That's what use_bayes 0 is for.
>> Unless you really have to, turning-off Bayes is a bad idea. The FP
>> rate
>> is much higher without it.
> Beyes does not make much sense in a multi-user, diverse community
> such as my university department. Makes sense here (small company;
> small user base)
bayes is usefull anythime, its just matter how its autolearn, trained
by users, setup, learned incorrect its waste of cpu time, and now bayes
is now a plugin so use_bayes 0 does not disable the plugin, just to
follow up on the thread :=)
Re: use_bayes=0 completly disables report function
Posted by John ffitch <jp...@codemist.co.uk>.
On Sat, 21 Apr 2012, RW wrote:
>
> That's what use_bayes 0 is for.
>
> Unless you really have to, turning-off Bayes is a bad idea. The FP rate
> is much higher without it.
>
Beyes does not make much sense in a multi-user, diverse community such as
my university department. Makes sense here (small company; small user
base)
==John ff
Re: use_bayes=0 completly disables report function
Posted by RW <rw...@googlemail.com>.
On Sat, 21 Apr 2012 14:27:42 +0200
Marcin Mirosław wrote:
> >>>> i've notice when i set use_bayes 0 then spamc -C report stops to
> >>>> work. I've got in log: spamd: Can't call method "learn" on an
> >>>> undefined value
> >>>
> >>> bayes_learn_during_report 0
> >>
> >> Hi darxus,
> >> thanks for reply. It solves my problem in half. I'd like to
> >> completly turn off bayes.
> >
> > Why does it only half solve your problem?
>
> Because option bayes_learn_during_report turns off bayes for reported
> messages but still bayes is used for classify incoming messages.
That's what use_bayes 0 is for.
Unless you really have to, turning-off Bayes is a bad idea. The FP rate
is much higher without it.
Re: use_bayes=0 completly disables report function
Posted by RW <rw...@googlemail.com>.
On Sat, 21 Apr 2012 07:38:32 +0200
Marcin Mirosław wrote:
> W dniu 2012-04-21 04:58, darxus@chaosreigns.com pisze:
> > On 04/20, Marcin Mirosław wrote:
> >> Hello,
> >> i've notice when i set use_bayes 0 then spamc -C report stops to
> >> work. I've got in log: spamd: Can't call method "learn" on an
> >> undefined value
> >
> > bayes_learn_during_report 0
>
> Hi darxus,
> thanks for reply. It solves my problem in half. I'd like to completly
> turn off bayes.
Why does it only half solve your problem?
Re: use_bayes=0 completly disables report function
Posted by Marcin Mirosław <ma...@mejor.pl>.
W dniu 2012-04-21 04:58, darxus@chaosreigns.com pisze:
> On 04/20, Marcin Mirosław wrote:
>> Hello,
>> i've notice when i set use_bayes 0 then spamc -C report stops to work.
>> I've got in log: spamd: Can't call method "learn" on an undefined value
>
> bayes_learn_during_report 0
Hi darxus,
thanks for reply. It solves my problem in half. I'd like to completly
turn off bayes. I'm guessing, as for now, it's the best to set
connection to non existent database (disadvantage is high number of
error messages in log).
Thanks,
Marcin
Re: use_bayes=0 completly disables report function
Posted by da...@chaosreigns.com.
On 04/20, Marcin Mirosław wrote:
> Hello,
> i've notice when i set use_bayes 0 then spamc -C report stops to work.
> I've got in log: spamd: Can't call method "learn" on an undefined value
bayes_learn_during_report 0
--
"Safe is anywhere a hungry person can't walk in three days." - John Titor
http://www.ChaosReigns.com
Re: use_bayes=0 completly disables report function
Posted by Marcin Mirosław <ma...@mejor.pl>.
W dniu 2012-04-21 15:29, Axb pisze:
Hello,
> http://spamassassin.apache.org/full/3.3.x/doc/Mail_SpamAssassin_Conf.txt
>
> You want:
>
> use_bayes 1
> use_bayes_rules 0
> use_learner 1
I found use_learner 0 is better for me, bayes engine isn't called while
message is reported.
Thanks for tip!
Marcin
Re: use_bayes=0 completly disables report function
Posted by Axb <ax...@gmail.com>.
On 04/20/2012 11:20 PM, Marcin Mirosław wrote:
> Hello,
> i've notice when i set use_bayes 0 then spamc -C report stops to work.
> I've got in log: spamd: Can't call method "learn" on an undefined value
> at /usr/lib64/perl5/vendor_perl/5.12.4/Mail/SpamAs
> sassin/PerMsgLearner.pm line 111,<GEN187> line 117.
>
> Whay i've to use bayes when i want to report spam to
> spamcop/dcc/pyzor/razor/etc ? Couldn't it be independent?
> Regards,
> Marcin
http://spamassassin.apache.org/full/3.3.x/doc/Mail_SpamAssassin_Conf.txt
You want:
use_bayes 1
use_bayes_rules 0
use_learner 1
h2h
Axb