http://issues.apache.org/SpamAssassin/show_bug.cgi?id=4787
Summary: BAYES_99 hits on all mail
Product: Spamassassin
Version: SVN Trunk (Latest Devel Version)
Platform: Other
OS/Version: other
Status: NEW
Severity: normal
Priority: P5
Component: Learner
AssignedTo: [email protected]
ReportedBy: [EMAIL PROTECTED]
Anyone ever seen anything like this? I'm wondering if there is a compelling
reason make SA stop learning spam tokens if the ham:spam token ratio exceeds a
certain level???
I guess I could increase the bayes_auto_learn_threshold_spam higher than 12 to
limit the amount of spam tokens learned... but it would be cool to have the
learner shut off spam token learning if the ratio is out of whack... or visa
versa with ham token learning if it outweighs the spam token count.
FWIW, this is the first time i've ever seen this happen since moving to SQL
bayes.
[EMAIL PROTECTED] spamassassin]# grep result: spamd.log | wc -l
12914
[EMAIL PROTECTED] spamassassin]# grep result: spamd.log | grep BAYES_99 | wc -l
12909
Sending a ham sample through hits like this...
X-Spam-Bayes-Tc-Spammy: 100
X-Spam-Status: No, hits=4.5 required=5.0
X-Spam-Bayes-Spammy-Tokens: 1.000-+--H*RU:rdns, 1.000-+--H*RU:helo,
1.000-+--H*RU:ident, 1.000-+--H*RU:intl, 1.000-+--H*RU:envfrom,
1.000-+--H*RU:auth, 1.000-+--HTo:D*net, 1.000-+--H*Ad:D*net,
1.000-+--H*F:D*com, 1.000-+--here
X-Spam-Bayes-Tc-Hammy:
X-Spam-Score: 4.5
X-Spam-Level: ****
X-Spam-Bayes-Hammy-Tokens:
X-Spam-Bayes: 1.0000
X-Spam-Bayes-Tc-Learned: 101
X-Spam-Bayes-Summary: Tokens: new, 53; hammy, 0; neutral, 1; spammy, 100.
X-Spam-Bayes-Tc: 154
X-Spam-Report: 4.5 points, 5.0 required
* 1.0 NO_REAL_NAME From: does not include a real name
* 3.5 BAYES_99 BODY: Bayesian spam probability is 99 to 100%
* [score: 1.0000]
mysql> select * from bayes_vars;
+----+----------+------------+-----------+-------------+-------------+----------
--------+--------------------+------------------+------------------+
| id | username | spam_count | ham_count | token_count | last_expire |
last_atime_delta | last_expire_reduce | oldest_token_age | newest_token_age |
+----+----------+------------+-----------+-------------+-------------+----------
--------+--------------------+------------------+------------------+
| 1 | $GLOBAL | 523306 | 67272 | 4680055 | 1139497034
| 7200 | 51609 | 1139314224 | 1139501076 |
+----+----------+------------+-----------+-------------+-------------+----------
--------+--------------------+------------------+------------------+
1 row in set (0.00 sec)
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.
