Re: sa-learn spam and Bayes_50

2009-10-28 Thread RW
On Tue, 27 Oct 2009 23:09:10 -0400 Alex wrote: > In either case, is there a way to exclude mails with USER_IN_WHITELIST > altogether? I have my ham level set at -0.3, but the USER_IN_WHITELIST > (and there are quite a few) adds -100.0, automatically making it ham. > I'm concerned that a spoofed m

Re: sa-learn spam and Bayes_50

2009-10-28 Thread Matus UHLAR - fantomas
> Benny Pedersen wrote: > > On tir 27 okt 2009 18:44:28 CET, John Hardin wrote > > > >>> 0.000 0 112532 0 non-token data: nspam > >>> 0.000 0844 0 non-token data: nham > > > > try to get them more equal numbered in your trains > > > >> reflect yo

Re: sa-learn spam and Bayes_50

2009-10-27 Thread Alex
Hi, > Instead, be more selective about the spam you train.  Only train > messages that completely missed with respect to Bayes (e.g. a spam > that got BAYES_00 or a ham that got BAYES_80) rather than corner cases > (e.g. a spam with BAYES_50 that got marked as spam, a spam that got > marked as BAY

Re: sa-learn spam and Bayes_50

2009-10-27 Thread Sam
John Hardin a écrit : On Tue, 27 Oct 2009, Sam wrote: John Hardin a écrit : On Tue, 27 Oct 2009, Sam wrote: > And after learning with sa-learn, it is still saying bayes_50 > whereas sa-learn told it has learned it. Okay, basic Bayes troubleshooting questions: (1) Are you running sa-l

Re: sa-learn spam and Bayes_50

2009-10-27 Thread Adam Katz
Benny Pedersen wrote: > On tir 27 okt 2009 18:44:28 CET, John Hardin wrote > >>> 0.000 0 112532 0 non-token data: nspam >>> 0.000 0844 0 non-token data: nham > > try to get them more equal numbered in your trains > >> reflect your actual raw spam

Re: sa-learn spam and Bayes_50

2009-10-27 Thread Benny Pedersen
On tir 27 okt 2009 18:44:28 CET, John Hardin wrote 0.000 0 112532 0 non-token data: nspam 0.000 0844 0 non-token data: nham try to get them more equal numbered in your trains reflect your actual raw spam/ham ratio, but yours is a little stro

Re: sa-learn spam and Bayes_50

2009-10-27 Thread John Hardin
On Tue, 27 Oct 2009, Sam wrote: John Hardin a écrit : On Tue, 27 Oct 2009, Sam wrote: > And after learning with sa-learn, it is still saying bayes_50 > whereas sa-learn told it has learned it. Okay, basic Bayes troubleshooting questions: (1) Are you running sa-learn as the same user t

Re: sa-learn spam and Bayes_50

2009-10-27 Thread Sam
Adam Katz a écrit : Sam wrote: I run spamassassin quite fine on a debian-lenny system. But I'm having a problem with sa-learn --spam and 1 message : But Bayes still show BAYES_50 : The Bayesian algorithm adds tokens from messages it is taught. These tokens are then added to the data

Re: sa-learn spam and Bayes_50

2009-10-27 Thread Sam
RW a écrit : On Tue, 27 Oct 2009 15:01:39 +0100 Sam wrote: RW a écrit : On Tue, 27 Oct 2009 13:33:14 +0100 If you find it surprising that that can happen, you don't understand how Bayes works. It's a leaning system that's intended to classify mail it hasn't seen based on

Re: sa-learn spam and Bayes_50

2009-10-27 Thread Sam
John Hardin a écrit : On Tue, 27 Oct 2009, Sam wrote: Oct 27 00:28:24 lenny spamd[20399]: spamd: clean message (0.0/5.0) for samueldu...@ingescom.com:102 in 4.8 seconds, 9803 bytes. Oct 27 00:28:24 lenny spamd[20399]: spamd: result: . 0 - BAYES_50,HTML_IMAGE_RATIO_08,HTML_MESSAGE,MISSING_MID

Re: sa-learn spam and Bayes_50

2009-10-27 Thread RW
On Tue, 27 Oct 2009 15:01:39 +0100 Sam wrote: > RW a écrit : > > On Tue, 27 Oct 2009 13:33:14 +0100 > > If you find it surprising that that can happen, you don't understand > > how Bayes works. It's a leaning system that's intended to classify > > mail it hasn't seen based on mail it has seen.

Re: sa-learn spam and Bayes_50

2009-10-27 Thread John Hardin
On Tue, 27 Oct 2009, Sam wrote: Oct 27 00:28:24 lenny spamd[20399]: spamd: clean message (0.0/5.0) for samueldu...@ingescom.com:102 in 4.8 seconds, 9803 bytes. Oct 27 00:28:24 lenny spamd[20399]: spamd: result: . 0 - BAYES_50,HTML_IMAGE_RATIO_08,HTML_MESSAGE,MISSING_MID scantime=4.8,size=9803,

Re: sa-learn spam and Bayes_50

2009-10-27 Thread Adam Katz
Sam wrote: >> I run spamassassin quite fine on a debian-lenny system. >> But I'm having a problem with sa-learn --spam and 1 message : >> But Bayes still show BAYES_50 : The Bayesian algorithm adds tokens from messages it is taught. These tokens are then added to the database's existing tokens an

Re: sa-learn spam and Bayes_50

2009-10-27 Thread Sam
Matus UHLAR - fantomas a écrit : On Tue, 27 Oct 2009 13:33:14 +0100 Sam wrote: I run spamassassin quite fine on a debian-lenny system. But I'm having a problem with sa-learn --spam and 1 message : ... But Bayes still show BAYES_50 : RW a écrit : If you find it surpr

Re: sa-learn spam and Bayes_50

2009-10-27 Thread Matus UHLAR - fantomas
>> On Tue, 27 Oct 2009 13:33:14 +0100 >> Sam wrote: >>> I run spamassassin quite fine on a debian-lenny system. >>> But I'm having a problem with sa-learn --spam and 1 message : >>> >>> ... >>> But Bayes still show BAYES_50 : > RW a écrit : >> If you find it surprising that that can happen, you d

Re: sa-learn spam and Bayes_50

2009-10-27 Thread Sam
RW a écrit : On Tue, 27 Oct 2009 13:33:14 +0100 Sam wrote: Hi, I run spamassassin quite fine on a debian-lenny system. But I'm having a problem with sa-learn --spam and 1 message : ... But Bayes still show BAYES_50 : If you find it surprising that that can happen, you don't underst

Re: sa-learn spam and Bayes_50

2009-10-27 Thread RW
On Tue, 27 Oct 2009 13:33:14 +0100 Sam wrote: > Hi, > > I run spamassassin quite fine on a debian-lenny system. > But I'm having a problem with sa-learn --spam and 1 message : > >... > But Bayes still show BAYES_50 : If you find it surprising that that can happen, you don't understand how Baye

sa-learn spam and Bayes_50

2009-10-27 Thread Sam
Hi, I run spamassassin quite fine on a debian-lenny system. But I'm having a problem with sa-learn --spam and 1 message : http://www.pastebin.org/48668 lenny:/home/samuel# sa-learn --dump magic 0.000 0 3 0 non-token data: bayes db version 0.000 0 112507