[Spambayes] PocketPC

Erin Lazzaro hera at optonline.net
Thu Feb 26 22:07:25 EST 2004

Thank you all for your responses.  I have read all the training pages in the Wiki, and I have devised a _Plan for Working with Pocket Inbox_.  I've only run one daily cycle so far - I'll report after a week on how good my filters are.  In the meantime, would anyone like to comment?

The goal is to get an effective filter relatively quickly, while spending as little time as possible on the desktop machine (since at-home time is precious, and my husband thinks I should play with him and not with my spam filters).

Training on mistakes and unsures seems the most intuitive, but since I have never yet had any ham classified as spam, I would expect the ratio to get very unbalanced.  Why do people think otherwise?  Do you start getting misclassified ham if the ratio gets too far out?

Initial setup:
	- Configure Outlook not to download mail automatically.
	- Configure ActiveSync to synchronize all three SpamBayes folders (Inbox, Junk, Junk Suspects)
	- Make two more mail folders, TraningHam and TrainingSpam (they don't need to be synchronized), and add the Spam field to their views.  Seed them with 5 ham and 5 spam (I just used the most recently arrived) and use SpamBayes Manger to train on them.
	- Add the Spam field to the Inbox view.

Each morning, on the train or elsewhere:
	- Review the three SpamBayes folders (Inbox, Junk, Junk Suspects) and move all mistakes and unsures to either TrainingHam or TrainingSpam.  (If I have time, I might also read and respond to the ham

Each night, on the desktop:
	- Sort TrainingHam and TrainingSpam on the Spam field and delete all but the 5 most out-of-whack (not considering duplicate spam and fake email bounces).
	- If there are fewer items in TrainingHam than in TrainingSpam, go into Inbox and copy the spammiest ham from there.  (If nothing in Inbox scores more than 0%, then let TrainingHam be short).
	- Train on TrainingHam and TrainingSpam, with the "rebuild" checkbox clear (this is the option I was missing before).  Clean the folders out afterwards.
-	Download the mail with the new filter.

Erin Lazzaro

More information about the Spambayes mailing list