[Spambayes] pre-filter the Review web page

amedee at amedee.be amedee at amedee.be
Tue May 9 19:30:25 CEST 2006

> How Can I pre-filter the training web page?  I get a lot of email from
> mailing lists that have a
> very low instance of SPAM and I would like to ignore them.  Is there a way
> I can have
> spambayes toss them so I can concentrate on the rest of the mail without
> having to sift
> through the mailing lists?  Is the only way to do this is to directly
> filter the data base?  If so
> what files do I need to touch with a bash script?

You didin't say how your mail is set up. I'll tell you how I solved it.
In my .procmailrc I move email from trusted mailing lists to separate
folders BEFORE the SpamBayes filtering kicks in, like this:

### These emails are not filtered by SpamBayes

## Mailing lists

# Pandora User Base mailing lists
* ^List-Id:[^<]+<\/[^.]+\.pub\.telenet\.be
        * MATCH ?? ^\/[^.]+

# PUB, not from mailing list
* ^From:.*pub\.telenet\.be

* ^Subject:.*\[smokeping

* ^List-id: campagne\.duimenvoorjongeren\.be

## Logs

* ^From: logcheck

* ^Subject:.*\[SmokeAlert

* ^Subject:.*\[SNORT

* ^Subject: Cron

### SpamBayes filtering starts here

:0 fw:hamlock
| sb_filter.py

# Messages that are so obviously spam that we should not train on them
* ^X-SpamBayes-Classification: spam; 1.00

# Messages that are spam but we might want to train on them
* ^X-SpamBayes-Classification: spam

# Unsure messages must be copied to the unsure folder for training
:0 c
* ^X-SpamBayes-Classification: unsure

# Ham that doesn't score 0.00 is eligible for training as well
:0 c
* ^X-SpamBayes-Classification: ham; 0.0[2-9]

:0 c
* ^X-SpamBayes-Classification: ham; 0.1[0-9]

### These mailing lists are filtered by SpamBayes

# Spambayes mailing list
* ^List-Id:.*spambayes\.python\.org

* ^Subject:.*\[Spambayes\]

Hope this helps,

Amedee Van Gasse
amedee at amedee.be

More information about the SpamBayes mailing list