MailWasher Server uses a statistical content analysis to filter out junk messages. To identify junk mail, a filter looks at the entire context of an email, not just the heading and subject line. It then compares the content of messages that are known to be spam with messages that are known to be legitimate, in order to build a database of words that is used to predict or identify the probability that a future email is spam or not. The filter learns to identify new spam the more it analyzes incoming messages and can train itself to identify new patterns of spam.
You can select to enable MailWasher Server to train the filter using junk messages that have been identified by FirstAlert! and legitimate messages, so that when you select to use statistical content message filtering, you are also filtering mail through your training specifications.
It is recommended that you train the filter before using it. You can select to only train the filter and not use it to filter your messages until you want to.
For more information about statistical content analysis, see: