On Sat, 2013-04-27 at 20:48 -0400, Greg Troxel wrote:
> I've pointed out a meta-issue, which is that SA's score adjustment
> algorithm is reasonable for tests where senders and the tests don't have
> a relationship, but IMHO SA should have standards for whitelists that
> accept money to be listed.
Many people have had issues with returnpath. They claim to address
issues, but most of my interactions with them have been unsatisfactory.
I've pointed out a meta-issue, which is that SA's score adjustment
algorithm is reasonable for tests where senders and the tests don't have
a relationship, b
On Sat, 2013-04-27 at 19:00 -0400, Joe Acquisto-j4 wrote:
> > > Very interesting. However, I don't see any BAYES_xx markings in the
> > > headers at all.
> > > I assumed that is because it is not scoring yet, due to low samples.
> > > Or some other reason.
> >
> > that could be the reason, othe
On Sat, 27 Apr 2013, Joe Acquisto-j4 wrote:
I don't want to know how to see the tokens, etc (I do, but already know how).
I was curious about this BAYES_xx thing, which I gather is something I should
see in a message header.
Yes, the BAYES_## are rules that would show up in the hit-rules list
On Sat, 27 Apr 2013, Alex wrote:
Hi,
To feed "ham" to bayes, should one only user mis-flagged mail, or may one
use unflagged (below 5) mail?
Expressed differently, can one feed "good" messages, "sa-learn --ham
path-to-ham " as one might feed missed spam, "sa-learn --spam path-to-spam"
You
>>> On 4/27/2013 at 11:17 AM, Benny Pedersen wrote:
> Joe Acquisto-j4 skrev den 2013-04-27 13:37:
>
>> Very interesting. However, I don't see any BAYES_xx markings in the
>> headers at all.
>
> how is you bayes setup ?
>
> what gives "sa-learn --dump magic" ?
>
>> I assumed that is because i
>>> On 4/27/2013 at 1:20 PM, John Hardin wrote:
> On Fri, 26 Apr 2013, Joe Acquisto-j4 wrote:
>
>> So, I could just feed a bunch of good mail, to --ham, and spam that is
> correctly marked
>> as spam as well as missed spam, to --spam?
>
> Correct; the important part is that what you train with
Hi,
To feed "ham" to bayes, should one only user mis-flagged mail, or may one
>> use unflagged (below 5) mail?
>>
>> Expressed differently, can one feed "good" messages, "sa-learn --ham
>> path-to-ham " as one might feed missed spam, "sa-learn --spam path-to-spam"
>>
>
> You can train hams that h
27.04.2013 23:15, Karsten Br�ckelmann kirjoitti:
> Point being, am I correct in assuming these numbers roughly reflect your
> ham/spam ratio?
>
>> > 0.000 0 28252 0 non-token data: nspam
>> > 0.000 0 187579 0 non-token data: nham
Yes. I want more spam,
On Sat, 2013-04-27 at 11:59 +0300, Jari Fredriksson wrote:
> 27.04.2013 04:54, Karsten Bräckelmann kirjoitti:
> > And it is good advice to keep the initial training corpora to a
> > ratio roughly assembling your ham/spam ratio, or maybe 1/1. (At this
> > point, we're approaching woodoo. Learning 10
On Sat, 27 Apr 2013, Niamh Holding wrote:
Hello John,
Saturday, April 27, 2013, 12:50:34 AM, you wrote:
JH> Simple rule: train any ham that doesn't hit BAYES_00.
???
What about ham that hits BAYES_00 and shows autolearn=no ?
If a ham hits BAYES_00 that means the Bayes system did a good job
On Fri, 26 Apr 2013, Joe Acquisto-j4 wrote:
So, I could just feed a bunch of good mail, to --ham, and spam that is
correctly marked
as spam as well as missed spam, to --spam?
Correct; the important part is that what you train with must be *correctly
classified* - training a ham as spam is no
27.04.2013 18:24, Benny Pedersen kirjoitti:
> Jari Fredriksson skrev den 2013-04-27 10:59:
>
>> 0.000 0 28252 0 non-token data: nspam
>> 0.000 0 187579 0 non-token data: nham
>>
>> I have no problems with Bayes whatsoever.
>
> this is an good working m
Niamh Holding skrev den 2013-04-27 18:25:
What about ham that hits BAYES_00 and shows autolearn=no ?
if its spam, sa-learn --spam else the above is ok, its no need to learn if it already is learned as
ham
--
senders that put my email into body content will deliver it to my own
trashcan, so
Hello John,
Saturday, April 27, 2013, 12:50:34 AM, you wrote:
JH> Simple rule: train any ham that doesn't hit BAYES_00.
???
What about ham that hits BAYES_00 and shows autolearn=no ?
--
Best regards,
Niamhmailto:ni...@fullbore.co.uk
pgp3P8oEu1ldu.pgp
Description
Jari Fredriksson skrev den 2013-04-27 10:59:
0.000 0 28252 0 non-token data: nspam
0.000 0 187579 0 non-token data: nham
I have no problems with Bayes whatsoever.
this is an good working mta setup, not a bayes problem :)
--
senders that put my e
Joe Acquisto-j4 skrev den 2013-04-27 13:37:
Very interesting. However, I don't see any BAYES_xx markings in the
headers at all.
how is you bayes setup ?
what gives "sa-learn --dump magic" ?
I assumed that is because it is not scoring yet, due to low samples.
Or some other reason.
that c
Joe Acquisto-j4 skrev den 2013-04-27 01:38:
path-to-ham " as one might feed missed spam, "sa-learn --spam
path-to-spam"
yes, but if you sort based on scores there is no point in using bayes
in the first place
only thing that is important is to feed what is spam and what is ham to
learning
Alain Kelder skrev den 2013-04-27 13:26:
RCVD_IN_RP_CERTIFIED=-3
RCVD_IN_RP_SAFE=-2
meta RP_FAIL_ADJ (RCVD_IN_RP_CERTIFIED && RCVD_IN_RP_SAFE)
score RP_FAIL_ADJ 5
just an example that works for me atleast, when i used dnseval rules
:=)
I've overridden the scores for the above tests in my
Axb skrev den 2013-04-27 09:06:
My (unrequested) advice: avoid Fedora! It's a Beta for the real
thing, and the real thing is either RHE or Centos
dont use any precompiled problems, best advice to give
--
senders that put my email into body content will deliver it to my own
trashcan, so if yo
Hi,
I'm experimenting with trying to get the latest v3.4.0 svn installed on an
>>> fc17 box, and it's complaining about a missing RabinKarpAccel.pm module.
>>>
>>> Can someone explain to me if this is needed in production, where I might
>>> find the latest version, and what exactly it does?
>>>
>>
Hi,
I'm experimenting with trying to get the latest v3.4.0 svn installed on an
>> fc17 box, and it's complaining about a missing RabinKarpAccel.pm module.
>>
>> Can someone explain to me if this is needed in production, where I might
>> find the latest version, and what exactly it does?
>>
>
> Goo
On 27.04.13 04:26, Alain Kelder wrote:
Today a spam message from livingsocial.com got through. It was sent
to a honey pot address (e.g. not used for legitimate mail). I don't
even have an account with livingsocial.com. This spam message would
have been caught had SA not credited it -5 points wi
On 27.04.13 09:50, Blason rock wrote:
Oh yeah. I rather should have posted my question in zimbra mailing list.
But any way i found the answer; its pretty easy to integrate prerbl with
postfix.
Yes, and it's not SA issue but the MTA's.
However, many users prefer only using SA and not any SMTP-ti
Do train those, which have a Bayesian probability close(r) to 0.5. Or
even worse, have a Bayesian probability contrary to the overall score,
or actual classification.
Training the plethora of spam hitting BAYES_99 might not be a mistake.
But it is pretty likely, to *not* improve general SA perfor
> Hello,
>
> Today a spam message from livingsocial.com got through. It was sent to
> a
> honey pot address (e.g. not used for legitimate mail). I don't even
> have
> an account with livingsocial.com. This spam message would have been
> caught had SA not credited it -5 points with two RP rules:
>
. . .
> Do train those, which have a Bayesian probability close(r) to 0.5. Or
> even worse, have a Bayesian probability contrary to the overall score,
> or actual classification.
>
> Training the plethora of spam hitting BAYES_99 might not be a mistake.
> But it is pretty likely, to *not* improve
27.04.2013 12:03, Axb kirjoitti:
> On 04/27/2013 10:59 AM, Jari Fredriksson wrote:
>> 27.04.2013 04:54, Karsten Bräckelmann kirjoitti:
>>> And it is good advice to keep the initial training corpora to a
>>> ratio roughly assembling your ham/spam ratio, or maybe 1/1. (At this
>>> point, we're approa
On 04/27/2013 10:59 AM, Jari Fredriksson wrote:
27.04.2013 04:54, Karsten Bräckelmann kirjoitti:
And it is good advice to keep the initial training corpora to a
ratio roughly assembling your ham/spam ratio, or maybe 1/1. (At this
point, we're approaching woodoo. Learning 10 times more ham than s
27.04.2013 04:54, Karsten Bräckelmann kirjoitti:
> And it is good advice to keep the initial training corpora to a
> ratio roughly assembling your ham/spam ratio, or maybe 1/1. (At this
> point, we're approaching woodoo. Learning 10 times more ham than spam is
> most likely to be a bad choice, thou
On 04/27/2013 06:09 AM, Alex wrote:
Hi all,
I'm experimenting with trying to get the latest v3.4.0 svn installed on an
fc17 box, and it's complaining about a missing RabinKarpAccel.pm module.
Can someone explain to me if this is needed in production, where I might
find the latest version, and w
31 matches
Mail list logo
