Details of Mailguard's pre-delivery spam evaluation process
Anti-spam protection and MailHurdle combine to use multiple methods for preventing spam from reaching the user's Inbox:
- Rejection of email from computers on a blocklist of IP addresses known for sending spam or on networks designated by ISPs for consumer use. For this, Mailguard uses a commercial and well-known repository of IP addresses referenced by many providers of anti-spam software;
- Evaluation of the sender domain 'SPF record' for disposition of email from IP addresses in their domain. SPF is commonly used and more strictly enforced by banks to prevent spoofing of email from their domain in 'phishing' attempts;
- Checking email against MailHurdle patterns of previous mail routing between people. When a new pattern is detected, the initial email is deferred but the retry and subsequent emails are accepted without delay. It is standard procedure for email servers to try resending a message many times, so the initial delay should be modest. This new feature favors email from the familiar and protects against 'hit-and-run' spam;
- Evaluating the permitted email content and assigning a spam score. Email scoring the maximum spam score of 300 are deleted; those scoring between 50 and 299 are deposited to the recipient's Junk Mail folder. Rules for spam scoring are done using a proprietary set of definitions and engine that is widely used in commercial anti-spam software.
Users can report erroneous suspect spam by viewing the messages in Webmail. This information is reported to the anti-spam vendor for refinement of anti-spam scoring techniques. It will not have an immediate impact on specific emails or senders. For that, use whitelists if necessary.