Train messages based on IP address

I’ve searched before I posted this and didn’t find anything.

I’d like to train incoming messages based on IP address. So if I receive an email from myself which is not from my IP address, it would be marked as spam. Conversely, any email from my server’s IP would not be filtered as junk.

Is this possible?

SpamSieve automatically learns from IP addresses when you train it with good and spam messages, but it does not accept or block messages solely on that basis. If you must do that, you can create whitelist or blocklist rules that match the Received header. However, my general advice is to correct all the mistakes and leave it at that. Micromanaging is rarely worthwhile.

The thought is that this would make the training for your own email address or mail from your own server better, only accepting emails from yourself as not spam if they contain your server’s IP and not some other spammer’s IP.

Yes, I understand the thought. It’s not clear to me that this would necessarily work, since it’s also possible for messages that you receive to contain your server’s IP.

I tend to be suspicious when people want to make manual rules because they can have unintended side effects and mask other problems. If you aren’t having problems with good messages going to the spam folder, why bother? And if you are, there’s probably some other problem that needs to be addressed.

Fair enough. I think if I have it trained better it will be fine. I think I haven’t had it setup as good as it should be to this point. Thanks for the quick responses.