teach the spam filter to learn from mistakes?
The spam filter can be taught that a message is spam, or not spam (aka ‘ham’) and will learn over time. Using Bayesian logic, the filter teaches itself when a message has a very low score that its contents tend to not be spam, and when a message has a very high score that its contents tend to be spam. This way, those messages which are in the middle can be pushed one way or another based on the content of the message and Bayesian inference. If a message was erroneously flagged as spam, or let through when it clearly was spam, you can help to teach the filter where it went wrong. There is a global folder on the mail server, called LearnSpam, where you can put those messages which were incorrectly classified. You don’t need to let us know about a message being there, as we can see them and shouldn’t have trouble figuring out if it was wrongly flagged or allowed to pass, unless you want us to explain why a message was flagged. In that case, please send us a message describing the mail you
