I'd like to train my Spam Assassin with some spam that it's missed, however I'm confused. With sa-learn I could target:

  • raw message
  • text-message (body)
  • html (body)

What one of those options is this best. I'd like to pull the message right out of my E-mail box. Is it of any concern if the raw-message has in the header's destination my e-mail address? Will that perhaps confuse its ability to learn spam from ham?

Any other suggestions about sa-learning


I use the raw message. There's no need to trim the messages before running sa-learn.

  • 1
    An advantage of using the raw message is that sa-learn can also process the headers, which may be more useful than the body itself. – John Gardeniers Jul 29 '12 at 0:44
  • As long as you give it separate --ham and --spam runs, it's all good. – Avery Payne Jul 3 '14 at 20:40

Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Not the answer you're looking for? Browse other questions tagged or ask your own question.