Report messages as Spam or Not Spam/Train Messages

If you find spam is not being detected by the filter or that legitimate messages are being caught as spam, the original messages should be uploaded to the SpamExperts accuracy training system. This allows the system to gather valuable information on the nature of the emails and helps to reduce future inaccuracies in message classification.

Messages trained must be the original email received by the recipient with the original message headers. Forwarded or copied message content is not sufficient for our system.

Emails that are configured to be unfiltered are excluded from the training system, for example (but not limited to), if a sender/recipient is on the Allow list, or if the mailbox is set to Filtered (incoming) = No in the Mailboxes Overview/Configuration page.

Messages with classifications outside of our control such as (but not limited to) remote host used invalid HELO/EHLO greeting can only be corrected by the sending server and training these will not make any changes to how our systems classify these messages in the future.

You can also use the following end-user client integrations with SpamExperts to report spam:

How to Train Messages

Training messages using this feature is available from the Domain Level Control Panel or the Email User Control Panel only.

To access the Train Messages page when logged in as an Admin Level User:

  1. In the General section, open the Domains Overview page
  2. Select the Domain with the misclassified message(s) to open the Domain Level Control Panel

Once you are viewing the affected Domain or User's Control Panel, proceed to train the message as Spam/Not Spam:

  1. Select Incoming > Train messages
  2. In the Train messages page, drag and drop or browse for and upload messages you want the system to treat differently. Messages can be in .msg, .eml or .txt format only

    3. The maximum size limit per message through the dashboard is 5 MB. For messages over this size, please use other methods of reporting spam or not spam.

  3. Select whether the message(s) should be recognized as legitimate (false positive) or spam (false negative)
  4. Click Continue
  5. You will be asked to confirm you wish to train a message, click Report as spam/Report as legitimate

  6. Close the confirmation box once you see the file(s) have been successfully uploaded

    7. /

The emails reported via all methods must be in .msg, .eml (IETF FRC 5322) or .txt format and contain the full unmodified message headers and body