SA Ignoring Bayes

[Spamassassin Users List <sa...@UniXpress.com>]

OS: SuSE 9.0
SA Version 2.55

I ran sa-learn on nearly 6,000 spam and 40,000 ham messages and rebuilt 
the database and exlicitly enabled bayesian checking but it still does not 
work.

I copied the global configuration file to ~.spamassassin and edited:

	# Add your own customisations to this file.  See 'man Mail::SpamAssassin::Conf'
	# for details of what can be tweaked.
	# 
	# rewrite the Subject: line with ****SPAM**** .* if set to 1 (default=1)
	rewrite_subject 0
	# report briefly, recommended for report_header==1 (default=0)
	use_terse_report 1

	auto_learn 1
	auto_learn_threshold_spam 10.0

	use_bayes 1
	bayes_ignore_header X-Bogosity

I ran spamassassin -D --lint:

	debug: Score set 0 chosen.
	debug: running in taint mode? no
	debug: ignore: using a test message to lint rules
	debug: using "/usr/share/spamassassin" for default rules dir
	debug: using "/etc/mail/spamassassin" for site rules dir
	debug: using "/home/lucky/.spamassassin" for user state dir
	debug: using "/home/lucky/.spamassassin/user_prefs" for user prefs file
	debug: using "/home/lucky/.spamassassin" for user state dir
	debug: bayes: 6110 tie-ing to DB file R/O /home/lucky/.spamassassin/bayes_toks
	debug: bayes: 6110 tie-ing to DB file R/O /home/lucky/.spamassassin/bayes_seen
	debug: Score set 3 chosen.
	debug: Initialising learner
	debug: is Net::DNS::Resolver available? yes
	debug: trying (3) slashdot.org...
	debug: looking up MX for 'slashdot.org'
	debug: MX for 'slashdot.org' exists? 1
	debug: MX lookup of slashdot.org succeeded => Dns available (set dns_available to hardcode)
	debug: is DNS available? 1
	debug: running header regexp tests; score so far=0
	debug: running body-text per-line regexp tests; score so far=1.5
	debug: bayes corpus size: nspam = 4612, nham = 38914
	debug: tokenize: header tokens for *F = "ignore@compiling.spamassassin.taint.org"
	debug: tokenize: header tokens for *m = " 1084286567 lint_rules "
	debug: cannot use bayes on this message; db not initialised yet
	debug: bayes: not scoring message, returning 0.5
	debug: bayes: 6110 untie-ing
	debug: bayes: 6110 untie-ing db_toks
	debug: bayes: 6110 untie-ing db_seen
	debug: Razor2 is not available
	debug: running raw-body-text per-line regexp tests; score so far=1.5
	debug: running uri tests; score so far=1.5
	debug: uri tests: Done uriRE
	debug: running full-text regexp tests; score so far=1.5
	debug: Razor2 is not available
	debug: Current PATH is: /home/lucky/bin:/usr/local/bin:/usr/bin:/usr/X11R6/bin:/bin:/usr/games:/opt/gnome/bin:/opt/kde3/bin
	debug: Pyzor is not available: pyzor not found
	debug: DCC is not available: dccproc not found
	debug: all '*From' addrs: ignore@compiling.spamassassin.taint.org
	debug: all '*To' addrs: 
	debug: running meta tests; score so far=1.5
	debug: is spam? score=1.5 required=5 tests=DATE_MISSING,NO_REAL_NAME
	debug: bayes: 6110 untie-ing


I see the line:
	debug: cannot use bayes on this message; db not initialised yet
but haven't found anything in the docs or FAQ that tells me how to 
initialize the db. (I have run "sa-learn --rebuild")

How can I make SA use Bayesian analysis for users with large enough 
samples to have populated their databases?

Thank you,
Lucky Leavell

Re: SA Ignoring Bayes

[Theo Van Dinter <fe...@kluge.net>]

On Wed, May 12, 2004 at 01:56:29PM -0400, Spamassassin Users List wrote:
> 	debug: cannot use bayes on this message; db not initialised yet
> but haven't found anything in the docs or FAQ that tells me how to 
> initialize the db. (I have run "sa-learn --rebuild")

Getting that message doing the compile/lint test isn't a big issue.
The message basically means that there weren't enough tokens found in the
message that were already in the DB, so it can't generate a score for you.

FYI: the message is changed in 3.0 to read:

 cannot use bayes on this message; not enough usable tokens found

since it's clearer than "not initialised"...

-- 
Randomly Generated Tagline:
"Do not meddle in the affairs of wizards, for you are crunchy and good
 with ketchup."                  - Unknown

Re: SA Ignoring Bayes

[Lucky Leavell <sa...@UniXpress.com>]

On Wed, 12 May 2004, John Hardin wrote:

> On Wed, 2004-05-12 at 11:51, Lucky Leavell wrote:
> > On Wed, 12 May 2004, John Hardin wrote:
> > 
> > > On Wed, 2004-05-12 at 10:56, Spamassassin Users List wrote:
> > > > 	debug: bayes: 6110 tie-ing to DB file R/O /home/lucky/.spamassassin/bayes_toks
> > > > 	debug: bayes: 6110 tie-ing to DB file R/O /home/lucky/.spamassassin/bayes_seen
> > > 
> > > Per some earlier comments, try "chmod 666" on your Bayes files...
> > >
> > I tried that (per your suggestion) and made the .spamassassin directory 
> > world writeable too ... same results.
> 
> Yeah. Some others were saying it needs write permissions is the only
> reason I mentioned it. I'm actually running it here R/O with no problems
> - I don't like the idea of auto-learn.
> 
Does it matter that I am still running the un-daemonized version:

	/usr/bin/perl /usr/bin/spamassassin -x

invoked via advosys filter.sh script?

Thank you,
Lucky