You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@spamassassin.apache.org by Praveen Kumar <pr...@gmail.com> on 2007/03/19 07:50:15 UTC
spamassassin 3.1.8 fine tuning
Hi ALL,
I've integrated Spamassassin 3.1.8 with SUN Java messaging. It's working
fine but
success rate of spam-detection is very less (around 20-25%).
How can i fine tune to get best results?
TIA,
My conf files below:
local.cf :-
# This is the right place to customize your installation of SpamAssassin.
#
# See 'perldoc Mail::SpamAssassin::Conf' for details of what can be
# tweaked.
#
# Only a small subset of options are listed below
#
###########################################################################
# Add *****SPAM***** to the Subject header of spam e-mails
#
# rewrite_header Subject *****SPAM*****
# Save spam messages as a message/rfc822 MIME attachment instead of
# modifying the original message (0: off, 2: use text/plain instead)
#
# report_safe 1
report_safe 1
# Set which networks or hosts are considered 'trusted' by your mail
# server (i.e. not spammers)
#
# trusted_networks 212.17.35.
# Set file-locking method (flock is not safe over NFS, but is faster)
#
# lock_method flock
# Set the threshold at which a message is considered spam (default: 5.0)
#
# required_score 5.0
required_score 5.0
# Use Bayesian classifier (default: 1)
#
# use_bayes 1
use_bayes 1
# Bayesian classifier auto-learning (default: 1)
#
# bayes_auto_learn 1
bayes_auto_learn 1
# Set headers which may provide inappropriate cues to the Bayesian
# classifier
#
# bayes_ignore_header X-Bogosity
# bayes_ignore_header X-Spam-Flag
# bayes_ignore_header X-Spam-Status
# Enable or disable network checks
skip_rbl_checks 0
use_razor2 1
use_dcc 1
use_pyzor 1
# Mail using languages used in these country codes will not be marked
# as being possibly spam in a foreign language.
ok_languages all
# Mail using locales used in these country codes will not be marked
# as being possibly spam in a foreign language.
ok_locales all
++++
i added the below lines to init.pre file:
loadplugin Mail::SpamAssassin::Plugin::Razor2
loadplugin Mail::SpamAssassin::Plugin::Pyzor
loadplugin Mail::SpamAssassin::Plugin::DCC
++++++++++
Re: spamassassin 3.1.8 fine tuning
Posted by Magnus Holmgren <ho...@lysator.liu.se>.
On Monday 19 March 2007 09:22, ram wrote:
> On Mon, 2007-03-19 at 12:20 +0530, Praveen Kumar wrote:
> > Hi ALL,
> >
> > I've integrated Spamassassin 3.1.8 with SUN Java messaging. It's
> > working fine but
> >
> > success rate of spam-detection is very less (around 20-25%).
> >
> > How can i fine tune to get best results?
> >
> > TIA,
>
> Use rules_du_jour. This will download cf files from rulesemporium
No, use sa-update instead. It can download cf files from rulesemporium as well
as the official rule updates.
--
Magnus Holmgren holmgren@lysator.liu.se
(No Cc of list mail needed, thanks)
Re: spamassassin 3.1.8 fine tuning
Posted by ram <ra...@netcore.co.in>.
On Mon, 2007-03-19 at 12:20 +0530, Praveen Kumar wrote:
> Hi ALL,
>
> I've integrated Spamassassin 3.1.8 with SUN Java messaging. It's
> working fine but
>
> success rate of spam-detection is very less (around 20-25%).
>
> How can i fine tune to get best results?
>
> TIA,
Use rules_du_jour. This will download cf files from rulesemporium
There are a number of cf files which come along with spamassassin but
they are fairly conservative when scoring mails.
Using cf files from rulesemporium you can improve your effeiciency.
Also PBL lists at your MTA to stop accepting spam rather than thrashing
them at spamassassin
You could then include spf and DK plugins into spamassassin
all that will gets near 90-95% ( for me but ofcourse YMMV :-) )
Bayes work well most of the times but for some spammers bayes is a way
to get in by including nonsensical junk in their mails. I would suggest
use bayes with scores reduced ( both positive and negative ) would be a
good idea
Thanks
Ram
Re: spamassassin 3.1.8 fine tuning
Posted by "Daryl C. W. O'Shea" <sp...@dostech.ca>.
Praveen Kumar wrote:
> Hi ALL,
>
> I've integrated Spamassassin 3.1.8 with SUN Java messaging. It's working
> fine but
>
> success rate of spam-detection is very less (around 20-25%).
>
> How can i fine tune to get best results?
A setup that is only catching 20-25% of spam usually has a config
problem. Make sure that spamd is not running with the -L option (local
test only, no network tests) and that if needed you've configured your
trusted_networks manually.
http://wiki.apache.org/spamassassin/TrustPath
Daryl
Re: spamassassin 3.1.8 fine tuning
Posted by John Andersen <js...@pen.homeip.net>.
On Sunday 18 March 2007, Praveen Kumar wrote:
> Txs John for your quick resp.
>
> Yes, i did installing of all the plugins thru CPAN
> do i need to set path of these .pm files anywhere in any of conf files so
> that spamassassing looks that specific path for plugins?
The plugins just use other software. Installing them will
NOT suffice if that other software does not exist or has not
been configured.
You've said you have the razor2 plugin for example.
So the next step is to find out if razor2 is a) actually
installed, and b) configured.
Start with man razor-admin
> Could you please help me in training bayes?
You have to keep messages that are real non-spam.
Chances are you have quite a few of those saved up
in various inboxes. You send these thru spamassassin
by using sa-learn
In a shell, type man sa-learn to find out how this is done.
As well, save up any spam that makes it thru. Send this
thru sa-learn with the --spam flag.
> Can somebody share the configuration file that is working quite fine?
Your configuration is not the problem here. You are expecting
instant results, but there is a bit of work left to do.
--
_____________________________________
John Andersen
Re: spamassassin 3.1.8 fine tuning
Posted by Praveen Kumar <pr...@gmail.com>.
Txs John for your quick resp.
Yes, i did installing of all the plugins thru CPAN
do i need to set path of these .pm files anywhere in any of conf files so
that spamassassing looks that specific path for plugins?
Could you please help me in training bayes?
Can somebody share the configuration file that is working quite fine?
Thanks
Prvnrk
On 3/19/07, John Andersen <js...@pen.homeip.net> wrote:
>
> On Sunday 18 March 2007, Praveen Kumar wrote:
> > use_razor2 1
>
> Did you actually install and configure razor?
> Having the plugin is not enough.
>
> Also, you have to train bayes with a couple hundred spams
> and hams (specifically telling it which is which) before it
> becomes effective.
>
> --
> _____________________________________
> John Andersen
>
>
Re: spamassassin 3.1.8 fine tuning
Posted by John Andersen <js...@pen.homeip.net>.
On Sunday 18 March 2007, Praveen Kumar wrote:
> use_razor2 1
Did you actually install and configure razor?
Having the plugin is not enough.
Also, you have to train bayes with a couple hundred spams
and hams (specifically telling it which is which) before it
becomes effective.
--
_____________________________________
John Andersen