Bayes training and header records

[Quentin Campbell]

I have followed advice given earlier on this list for collecting
spam/ham to use for some additional training of the Bayes rules. In
particular I am keen to train SA to recognise the "false negatives" that
are currently getting through.

I am curious to know why it's considered necessary to remove the
"X-Newcastle-MailScanner..." headers but not the (X-)Received: headers
and the various "ReSent-..." headers that are added when I bounce
messages to my spam/ham mailboxes?

Specifically which headers should be removed and why?

What role do headers generally play in the Bayes scoring of spam and
non-spam messages?

Quentin
---
PHONE: +44 191 222 8209    Computing Service, University of Newcastle
FAX:   +44 191 222 8765    Newcastle upon Tyne, United Kingdom, NE1 7RU.
------------------------------------------------------------------------
"Any opinion expressed above is mine. The University can get its own."