Every email you receive is analysed by our SpamAssassin filter and given a score based on the likelihood of being spam, as follows:
- Messages with a score of 1 or above are considered to look at least a little bit spam-like, the higher the number the more spam-like it appears to be.
- Messages with a score of 0 do not seem particularly spam-like or ham-like (ham being a term for "not spam").
- Messages with a score of -1 or below are considered to look at least a little bit ham-like (so not spam-like). Again, the larger the negative number, the more ham-like it appears to be.
How to train the spam filter
To help improve the spam filter you can teach the learning AI which messages you consider to be spam. Simply open the message in Usermin's simplified inbox view, scroll to the bottom and click the "Report Spam" button. The learning AI in the spam filter will analyse the message and in future the system will move messages with similar traits to your spam folder automatically so you don't keep seeing the same messages.
Conversely, should a message be filtered into your spam folder that you don't consider spam, go to Mail > Spam in the top left corner to access your spam folder. Open the message, scroll to the bottom and click the "Report Ham" button to have the learning AI train itself to deliver similar messages into your inbox.
Training is a gradual process, so you may not see any changes in behaviour until you've reported a few emails as either spam or ham.
As most people don't use Usermin as their main email client, it can be helpful to create a folder to drag spammy messages into, and then every so often (eg once a week) you can just login to Usermin, open the folder, select all messages in the folder, and then hit the "Report Spam" button.
If you have any questions about managing the spam settings for your account, you can create a support ticket -- we're always happy to provide assistance.