BAYES_00 and FN

[Robert Fitzpatrick <li...@webtent.net>]

I see a lot of messages hitting BAYES_00 and reducing enough to make it
a FN. After some learning, problem solved, but still an issue for new
message types. Is there a way to protect from this sort of thing? Like a
recipe not to add the bayes score if the score is over 7 and BAYES_50 or
lower? Would that be detrimental to my scoring?

Thanks in advance!

-- 
Robert

Re: BAYES_00 and FN

[Matus UHLAR - fantomas <uh...@fantomas.sk>]

> On 22.02.08 14:04, Robert Fitzpatrick wrote:
> > I see a lot of messages hitting BAYES_00 and reducing enough to make it
> > a FN. After some learning, problem solved, but still an issue for new
> > message types. Is there a way to protect from this sort of thing? Like a
> > recipe not to add the bayes score if the score is over 7 and BAYES_50 or
> > lower?

On 23.02.08 22:40, Matus UHLAR - fantomas wrote:
> no, the BAYES_00 has to indicate that message is probbably not a spam,
> whatever other rules say. I've seen false positives with BAYES_00 and
> resulting score over 7.
> 
> > Would that be detrimental to my scoring?
> 
> yes.

and I forgot to add: train your spam database properly. Especially on false
positives/negatives.
-- 
Matus UHLAR - fantomas, uhlar@fantomas.sk ; http://www.fantomas.sk/
Warning: I wish NOT to receive e-mail advertising to this address.
Varovanie: na tuto adresu chcem NEDOSTAVAT akukolvek reklamnu postu.
Boost your system's speed by 500% - DEL C:\WINDOWS\*.*

Re: BAYES_00 and FN

[Matus UHLAR - fantomas <uh...@fantomas.sk>]

On 22.02.08 14:04, Robert Fitzpatrick wrote:
> I see a lot of messages hitting BAYES_00 and reducing enough to make it
> a FN. After some learning, problem solved, but still an issue for new
> message types. Is there a way to protect from this sort of thing? Like a
> recipe not to add the bayes score if the score is over 7 and BAYES_50 or
> lower?

no, the BAYES_00 has to indicate that message is probbably not a spam,
whatever other rules say. I've seen false positives with BAYES_00 and
resulting score over 7.

> Would that be detrimental to my scoring?

yes.

-- 
Matus UHLAR - fantomas, uhlar@fantomas.sk ; http://www.fantomas.sk/
Warning: I wish NOT to receive e-mail advertising to this address.
Varovanie: na tuto adresu chcem NEDOSTAVAT akukolvek reklamnu postu.
Micro$oft random number generator: 0, 0, 0, 4.33e+67, 0, 0, 0...