To make sure you can still use your e-mail program even if you receive a flood of advertising messages, you can have your incoming messages scanned automatically for the typical attributes of unwanted spam messages. Spam messages that are detected are handled based on how the spamscanner is configured and the results it returns.
To determine the possibilities and limitations of the spamscanner, it is helpful to familiarize yourself more closely with how it works. The spamscanner is a piece of software. This takes us straight to the most important thing to remember, namely that software is only as good as the idea behind it. This is especially important in the case of difficult tasks such as interpreting the content of e-mails. The rules and their priorities are designed to minimize errors in the form of undetected spam e-mails (false negatives). In some rare cases, it is still possible that “good” e-mails (ham) are incorrectly classified as spam (false positives). An e-mail message is scanned according to various rules. For each test that returns a positive result, the message is given a certain number of points that are added together. If the total number of points exceeds a specific critical value once all tests are complete, the message is treated as spam.
Some examples of these spam tests:
Is the subject of the e-mail written in all caps?
Does the e-mail contain no text and simply a picture attachment?
Does the text of the message contain known spam URLs?
Are parts of the e-mail header obviously fraudulent?
If only one or two tests is passed, the message usually does not reach the necessary number of points to be identified as spam. This prevents a message from being classified as spam if its subject was inadvertently written in all caps, for instance.
Our spamscanner uses the open source project SpamAssassin, which offers many configuration options. You can set several of these yourself in the Control Panel.
You can adjust the threshold that triggers a message to be classified as spam. The lower the value, the more sensitive the spamscanner will be. However, this also increases the risk that legitimate e-mails will be identified as spam. Conversely, if you increase the threshold, the risk of false positives is less, but the recognition rate of actual spam will decrease as well. The default value for required_score is 4 spam points.
If a message is identified as spam, you can have the subject flagged accordingly. This will add a user-defined string to the existing subject. By default, the string *****SPAM***** is added to the subject. You can also use the following keywords:
placeholder for the number of spam points counted
placeholder for the current threshold value
If you frequently receive undesired (advertising) e-mails from specific addresses but they are not identified by the spamscanner, you can add the sender to your own custom blacklist. You can enter full e-mail addresses or use the wildcards ‘*’ and ‘?’:
All e-mails from email@example.com will be treated as spam.
All e-mails from any sender with @example.com in the address will be treated as spam.
All e-mails from e-mail addresses that fit the pattern [letter].[letter]@example.com will be treated as spam.
The same principle can be used to add senders to your whitelist. E-mails from these addresses will never be classified as spam, regardless of the scores returned by the spam tests.
If you have enabled the spamscanner, you can activate the Spambox feature in the Control Panel. E-mails that are identified as spam will not land in your Inbox but, instead, in a subfolder called Spam. You can use Hostpoint Exchange to review the contents of your Spam folder or, alternatively, access this folder directly via IMAP. The e-mails in the Spam folder are deleted automatically after a user-defined period of one to seven days in order to clear space for new messages.
For more information about SpamAssasin, please refer to
the information on the topic of undesired advertising e-mails (spam) provided by the Federal Data Protection and Information Commissioner (in german).