Chris Santerre wrote:
>> I thought I remembered a few years back that Baysian chains had a
>> 10% increase in capture rate over straight Bayes rules. I would
>> think that this is similar.
Marc Perkel wrote:
> I've always thought that a second basian filter that would just
> look at rule hits woul
On Mon, 12 Oct 2009 10:49:06 -0700
Ted Mittelstaedt wrote:
> I think if you sit down and start trying to define examples
> and run them through large databases of spam and ham you
> will find that it doesen't work the way you think it does. That
> is what I was talking about when I said that st
Title: RE: SA needs a new paradigm for rule structure
Chris Santerre wrote:
> -Original Message-
> From: Ted Mittelstaedt [mailto:t...@ipinc.net]
> Sent: 2009-10-10 02:40
> To: Marc Perkel
> Cc: users@spamassassin.apache.org
> -Original Message-
> From: Ted Mittelstaedt [mailto:t...@ipinc.net]
> Sent: 2009-10-10 02:40
> To: Marc Perkel
> Cc: users@spamassassin.apache.org
> Subject: Re: SA needs a new paradigm for rule structure
>
>
> Marc Perkel wrote:
> > I've brough
RW wrote:
On Fri, 09 Oct 2009 23:40:01 -0700
Ted Mittelstaedt wrote:
I know that it seems like the idea of building up "meta" rules with
a lot of small rules will give you a more accurate hit rate, but
this is one of those non-intuitive things that can be shown by
statistical mathmatics, that
4 out 5 bad words fround, surely a 419 scam
scoreRAISEFLAG 100
__GOD does not score, __NIGERIA neither, etc, 4 out of 5 does, a 100 a per
your request.
--
View this message in context:
http://www.nabble.com/SA-needs-a-new-paradigm-for-rule-structure-tp25822909p25838064.html
Sent from the Spa
RW wrote:
On Fri, 09 Oct 2009 23:40:01 -0700
Ted Mittelstaedt wrote:
I know that it seems like the idea of building up "meta" rules with
a lot of small rules will give you a more accurate hit rate, but
this is one of those non-intuitive things that can be shown by
statistical
On Fri, 09 Oct 2009 23:40:01 -0700
Ted Mittelstaedt wrote:
>
> I know that it seems like the idea of building up "meta" rules with
> a lot of small rules will give you a more accurate hit rate, but
> this is one of those non-intuitive things that can be shown by
> statistical mathmatics, that i
Ted Mittelstaedt wrote:
Marc Perkel wrote:
I've brought this idea up over the years but I'll try to explain it
in a different way. Maybe we can do this with a lot of meta rules.
What we need are rules that combine a lot of simple rules into
concepts and then combine those rules into rules t
Marc Perkel wrote:
I've brought this idea up over the years but I'll try to explain it in a
different way. Maybe we can do this with a lot of meta rules.
What we need are rules that combine a lot of simple rules into concepts
and then combine those rules into rules that score - and score big.
On Fri, 9 Oct 2009, John Hardin wrote:
... it could trivially be done right now based on the existing evolver
if you simply fed it _all_ of the existing rules to use as its base, and
(for example) kept every evolved rule set whose fitness was > 10 (or
whatever turns up as a good cutoff poi
On Fri, 2009-10-09 at 15:31 -0400, MySQL Student wrote:
> I sent this message more than an hour ago, and it looks like it's yet
> to hit the list. Resending.
Indeed -- there was an issue with athene for a short period rejecting
valid mailing list posts. Got my previous reply to this thread around
Hi,
I sent this message more than an hour ago, and it looks like it's yet
to hit the list. Resending.
Thanks,
Alex
-- Forwarded message --
From: MySQL Student
Date: Fri, Oct 9, 2009 at 2:34 PM
Subject: Re: SA needs a new paradigm for rule structure
To: SA Mailing list
On Fri, 2009-10-09 at 11:34 -0700, John Hardin wrote:
> On Fri, 9 Oct 2009, Karsten Bräckelmann wrote:
> > Whoa, dude! You just left the heavy sarcasm in, and snipped everything
> > from the quote that clarifies this statement and identifies it as
> > sarcasm.
>
> I suspect that Alex was respondi
On Fri, 2009-10-09 at 11:40 -0700, Marc Perkel wrote:
> Karsten Bräckelmann wrote:
> > Maybe you really should read up on some docs, and actually have a look
> > at the stock rules, as well as some third-party rule-sets. Just to see
> > your "innovative" concept already in use. And maybe even unde
On Fri, 9 Oct 2009, Karsten Br?ckelmann wrote:
On Fri, 2009-10-09 at 13:28 -0400, "Alex" / "MySQL Student" wrote:
What we need are rules that combine a lot of simple rules into concepts
and then combine those rules into rules that score - and score big. As
an example, [...]
Yes, SA definitel
Karsten Bräckelmann wrote:
On Fri, 2009-10-09 at 08:14 -0700, Marc Perkel wrote:
I've brought this idea up over the years but I'll try to explain it in a
different way. Maybe we can do this with a lot of meta rules.
What we need are rules that combine a lot of simple rules into
On Fri, 2009-10-09 at 13:28 -0400, "Alex" / "MySQL Student" wrote:
> > > What we need are rules that combine a lot of simple rules into concepts
> > > and then combine those rules into rules that score - and score big. As
> > > an example, [...]
> >
> > Yes, SA definitely needs that and sorely lac
On Fri, 9 Oct 2009, Marc Perkel wrote:
What we need are rules that combine a lot of simple rules into concepts and
then combine those rules into rules that score - and score big. As an
example, lets take a standard nigerian scam email.
From <> reply to:
[I don't know you] Dear stranger, I am
Hi,
>> What we need are rules that combine a lot of simple rules into concepts
>> and then combine those rules into rules that score - and score big. As
>> an example, [...]
>
> Yes, SA definitely needs that and sorely lacks this ultimate feature!
Can I respectfully add to this that John Hardin h
On Fri, 2009-10-09 at 08:14 -0700, Marc Perkel wrote:
> I've brought this idea up over the years but I'll try to explain it in a
> different way. Maybe we can do this with a lot of meta rules.
>
> What we need are rules that combine a lot of simple rules into concepts
> and then combine those ru
I've brought this idea up over the years but I'll try to explain it in a
different way. Maybe we can do this with a lot of meta rules.
What we need are rules that combine a lot of simple rules into concepts
and then combine those rules into rules that score - and score big. As
an example, lets
22 matches
Mail list logo