A Smarter Spam Stopper
The SpamButcher spam stopper was developed in response to first-generation spam blocker software
such as McAfee's SpamKiller.
Typically these products suffer from poor spam mail catching performance in addition to intercepting legitimate email due to their simple text-based filtering.
With less sophisticated spam stoppers, a single unwanted word in a message can cause it to be filtered. Conversely, if spammers avoid using any of the phrases in the built-in filter, it will be rendered ineffective.
To address these short comings, SpamButcher introduced fuzzy logic spam filtering.
Fuzzy logic allows SpamButcher to evaluate subjective factors about the message collectively to determine if it's spam or not. A single offending phrase in a message from a friend will not trigger the spam stoppers. There is no publicly available list of filtered phrases that spammers can easily avoid.
This lets the filter target phenomena that occur in both spam and normal email. Many phrases that appear in a large portion of spam, still periodically show up in email from a friend of colleague. SpamButcher is capable of considering this information in comparison to other factors within each message.
Different users of the spam stopper obviously report different success levels. Commonly users report that SpamButcher stops in excess of 96% of their unwanted messages. This is enough to in effect eliminate their problem with junk email.