http://www.ritlabs.com New posts in How to use Bayesit! of forum at www.ritlabs.com [www.ritlabs.com] en http://backend.userland.com/rss2 Tue, 31 Jan 2006 15:17:23 +0200 How to use Bayesit! in forum The Bat! - Configuring the E-mail Client.
Thanks nlkes!

In the mean time I have given up.
So I created a dummy adress at my own mail server and forward gmail there.
Then I can use K9 again, a big relief...



31 January 2006 15:17:23, bat99.]]> http://www.ritlabs.com/en/forums/forum4/topic2483/message10116/ http://www.ritlabs.com/en/forums/forum4/topic2483/message10116/ Tue, 31 Jan 2006 15:17:23 +0200 The Bat! - Configuring the E-mail Client How to use Bayesit! in forum The Bat! - Configuring the E-mail Client.
you can find some newer information on http://www.ritlabs.com/en/solutions/BayesIt.php (this URL ist given by the plug-in itself - use the context menu entry "information" on the plug-in entry in the preferences dialog) - however, these pages are not totally up to date, too...

to my mind one of the most common beginning problems when using BayesIt is not training non-spam! give the filter at least double the amount of non-spam training than spam training to avoid false positives... then you can safely use a lower spam threshold (I work with 60%)

by the way: 0.8.1 is NOT current... there's at least 0.8.4
31 January 2006 09:57:37, nlkes.]]> http://www.ritlabs.com/en/forums/forum4/topic2483/message10109/ http://www.ritlabs.com/en/forums/forum4/topic2483/message10109/ Tue, 31 Jan 2006 09:57:37 +0200 The Bat! - Configuring the E-mail Client How to use Bayesit! in forum The Bat! - Configuring the E-mail Client.
Does it exists any updated documentation on this?
I have seen the docs at http://klirik.narod.ru/usefuls/bayesiteng.htm, but it's   for rel. 04, current is 0.8.1.
This document describes an installation wizard  which does not appear with the current release.

My installation doesn't seem to work at all, if I set the limit to 98% all mail goes into spam.
If I set it to 99% none goes to spam, even if I have marked an identical mail as spam earlier...

After using K9 for some years I feel in the dark here. With K9 it's possible to see how mails are rated, cant find anything like this in Bayesit.
(Cant use K9 since I'm using gmail now)
24 January 2006 08:28:52, bat99.]]> http://www.ritlabs.com/en/forums/forum4/topic2483/message9985/ http://www.ritlabs.com/en/forums/forum4/topic2483/message9985/ Tue, 24 Jan 2006 08:28:52 +0200 The Bat! - Configuring the E-mail Client