Connect. Communicate. Collaborate. Securely.

Home » Kerio User Forums » Kerio Connect » Ham learning not being updated in Bayes DB
  •  
aj08

Messages: 57
Karma: 1
Send a private message to this user
I'm seeing an issue on Kerio Connect 8.0 on Mac servers where if a user marks a message as spam and then clicks the not spam button, the bayes statistics do not update for ham. I've used to used this method to quickly train the bayes db with messages that should be ham but that may score too high to be automatically learned as such. I've also relied on this method to quickly get bayes up to the required 200 messages learned as spam and ham before it will be used to modify scores.

If a message does score low enough, it is being learned as ham and the statistics are updating to reflect that. It's user trained ham that appears to not get learned. I've tried using the Kerio web interface and have also tried moving messages from the Junk mail box. In each case, the spam log shows that the message has been requested to be learned as ham but the numbers shown under bayesian filter statistics do not change.

I'm also seeing the following although I'm not certain it indicates why Kerio isn't learning my messages as ham.

Quote:
4/Feb/2013 19:21:16][2955980800] {spamassassin} Perl_stderr: [32004] dbg: bayes: Using userid: 1
[14/Feb/2013 19:21:16][2955980800] {spamassassin} Perl_stderr: [32004] dbg: bayes: =====> _wait_for_write_lock
[14/Feb/2013 19:21:16][2955980800] {spamassassin} Perl_stderr: [32004] dbg: bayes: ebd21c4c0c5aa194852ad09a476c2794e8d7af93@sa_generated already learnt as opposite, forgetting first
[14/Feb/2013 19:21:16][2955980800] {spamassassin} Perl_stderr: [32004] dbg: check: pms new, time limit in 299.985 s
[14/Feb/2013 19:21:16][2955980800] {spamassassin} Perl_stderr: [32004] dbg: bayes: =====> tie_db_writable
[14/Feb/2013 19:21:16][2955980800] {spamassassin} Perl_stderr: [32004] dbg: bayes: =====> _wait_for_write_lock
[14/Feb/2013 19:21:16][2955980800] {spamassassin} Perl_stderr: [32004] dbg: bayes: forget() returned a fatal error, so learn() will too
  •  
freakinvibe

Messages: 1540
Karma: 62
Send a private message to this user
If you have a message in your inbox, then you mark it as Spam, then you mark it as Ham, it will not count towards your score. As I see it, it will just remove the learned entry you made before.

For example if your statistics were

Spam=15, Ham=15

Then you get a mail in your inbox which does not count towards Ham (because the score was too high), your statistics are still

Spam=15, Ham=15

Now you mark that message as Spam. Your score will be

Spam=16, Ham=15

Now you mark the same message as Ham, your score will be

Spam=15, Ham=15

It will deduct the learned Spam, but not increase Ham.


Dexion AG - The Blackberry Specialists in Switzerland
https://dexionag.ch
Previous Topic: German: Einzelne Konten auf neuen Mailserver umziehen
Next Topic: Connect client 8
Goto Forum:
  


Disclaimer:
Kerio discussion forums are intended for open communication between forum members and may contain information and material posted by members which may be useful in learning about Kerio products. The discussion forums are not intended to provide technical support for any specific product. Any information implied or expressed in the discussion forums is that of the posting member. Kerio is in no way responsible for the information posted in the forums, or its accuracy. Kerio employees may participate in the discussions, but their postings do not represent an offical position of the company on any issues raised or discussed. Kerio reserves the right to monitor and maintain the forums to promote free and accurate exchange of information.

Current Time: Wed Oct 18 07:34:35 CEST 2017

Total time taken to generate the page: 0.00421 seconds
.:: Contact :: Home ::.
Powered by: FUDforum 3.0.4.