How is spam identified

  Every mail at is evaluated for spam and the result is saved in form of invisible part of the mail, called header:
X-Spam-Level: +++++++
  The amount of plus signs represents the probability that a mail is spam - the more pluses, the bigger the chance. Technically, the evaluation is done using Spamassassin 3.1 and Spamass-milt as glue between spamassassin and sendmail 8.13.5.

How to use the results

  For normal e-mail addresses, you can only adjust your spam filter. Later you can sort e-mail in your mail program using the results of the evaluation. It is not possible to reject the mails without delivering it.
  Based on the amount of accumulated pluses, the message for a mailing lists can be rejected, if the probability is too high. There are only two steps to adjust your mailing list to get less spam:

When is the evaluation not correct

   The big problem in the spam fight is, that there is no way to say for a mail if it is for sure spam or not. In order to achieve good results a lot of algorithms are used. Some of them are proceeded on the mail server itself, but the most data comes from the internet, where information about standard spam messages, is continiously updated. However for newly invented spam messages, there is no information yet, so they will have only few pluses.
  In other cases, when the mail looks according to the algorithms as spam, it will be marked as such, even if it isn't. However, this are seldom cases.

Final notes

  You are welcome to share your ideas about how the spam or ham is managed and if you have any suggestions for improvements, we will be more than happy to share them at A way to contribute is to cooperate for improving our spamtrap - mail address, which gets only spam. The idea is that you subscribe that e-mail to all possible spammers and once a mail is received to the spamtrap it will be remembered. All quite similar mails afterwards send to any address in will get higher probability.