The spam filter is a filter that tries to distinguish as much unwanted e-mail (spam) as possible from desired e-mail.

How it works

All incoming e-mail messages are reviewed. The filter looks at characteristic spam features such as language use, formatting and keywords such as "viagra", "bitcoin" etc. The more characteristics found, the more points an e-mail message gets. And the more points, the more likely a message is an unwanted (spam) message.

Spam filter for mailboxes

E-mail that the filter recognises as spam is placed in a separate ‘spam’ folder. You can also train the filter to detect spam even better. You do this by moving e-mail to the spam folder. If e-mail is wrongly placed in the spam box, you can mark it as not spam (ham) by moving it to another folder, e.g. your inbox. Only when you move mail from the spam box to the trash, it will not be marked as “not spam” . Via the webmail, it is also possible to report via the spam button:

Spam filter for Relay service ( own mail server)

Users of the relay service can also use the spam filter. By default, the filter will let all mail through, but will mark it in the headers. This is done using the following header: In case of (possible) spam X-Spamd-Bar: +++++++ X-Spam-Status: Yes, score=7.00 And no spam: X-Spamd-Bar: -- X-Spam-Status: No, score=-2.30 Based on these headers, action can then be taken on this mail in the own mail server. By default, from a score of 6 and above, the X-Spam-Status will mark mail with “Yes”. If a lot of spam is reported, this score can be adjusted downwards and visa versa. Based on the number of + signs or the number behind the X-Spam-Status, it is of course possible to take action earlier.