Minimizing False Positives
What is a False Positive?
As with any spam product, there is a chance that IMail Server may identify non-spam messages as spam. Such mistakes are called false positives. False positives can occur in both connection and content filtering. Example
Why Do False Positives Occur?
False positives resulting from content filtering may include newsletters and other various types of e-mail that people subscribe to. Many of these get caught by the content filters because they may contain ads that look like spam.
How to Prevent False Positives
The following methods are effective in minimizing false positives:
- White List (trusted addresses). Add the IP address (or range of addresses), domain names, and e-mail addresses for your network to the trusted address list. Any e-mail received from an IP address in this list will not have connection or content filtering performed on it.
- Set up delivery rules to send spam to a sub-mailbox in each user's directory. To do this, configure connection filtering and content filtering (Phrase Filter and the Statistical Filter) to place X-headers in the message. Then, create a domain rule that searches for a header that contains X- IMAIL-SPAM to place the message in a sub-mailbox (i.e. H~X-IMAIL-SPAM:spam). Users can then make individual rules that move messages back into the main mailbox.
- When a client connects to SMTPD32 and authenticates, incoming e-mail is not automatically checked by connection filtering. If the option on the Setting Domain Level Antispam Options page is not selected, mail from authenticated users will not undergo content filtering.