Q: How does the new “mark as spam” feature work?
A: As we announced in our last newsletter the SpamTitan Outlook Add-in is now available but what exactly happens if you use the Outlook add in to mark something as SPAM?
When you send mail flagged as Spam, the email is passed to your local SpamTitan to be learned as spam. You are sending the email message body to be trained by the SpamTitan Bayesian filter as a token, aka either Spam or Ham.
You can view trained tokens under “Anti-Spam Engine ->Settings-> Bayes Database” (keep in mind that other services feed into this count). You can see trained emails under “Logs->Interface->Cfma.log” and search for outlook. The Bayes filtering applies to all domains.
Q: What to do if you receive false negatives (meaning spam that passes filtering)?
A: Firstly the SpamTitan anti spam filter has an extremely high spam catch rate and very low false positive rate. (link to VB). If an occasional spam passes the filter the Outlook Addin allows you to submit false negatives to SpamTitan’s Bayesian Database (powered by SpamAssassin) for training over HTTP.
SpamTitan will learn the message bodies of submitted emails, searching for common terms in the mail and learning those as tokens. A token can be either Spam or Ham (legitimate mail), if this token is found in future emails then the spam score for the mail will rise, thus preventing future false negatives.
If you have any questions or would like more information please contact one our engineers at spamtitan@titanhq.com.