Death2Spam

Its third-generation spam detection technology includes academically-proven statistical analysis of the word-frequency (probability) distributions from vast numbers of good and spam email messages. Death2Spam checks every email against hundreds of discriminators.

Death2Spam compares each piece of email against:

• huge databases of known spam, collected from Internet mailboxes over many years.
• how you've taught it to classify your email in your individual or organizational preferences.
• learning from other users - the Death2Spam global hive learns every time a user classifies an email as spam or otherwise.

 

The Death2Spam "classifier" is an Adaptive Expert System, which is able to accurately assign each email a probability score representing the likelihood that it's spam. A probability score of 100 means the email is certainly spam. A probability of zero means it's as good as it gets.

If there are too few clues in the message (very rare), or if it contains an equal number of good and spam words, the probability score will hover somewhere around 50, in a fuzzy zone where the classifier is"unsure". Users can classify their personal "unsure" emails good or bad and Death2Spam will remember the individual classifications.

It is one of Death2Spam's strengths that the unsure zone's boundaries have been fine-tuned to reduce false-positive classifications to very low levels, while keeping the total number of uncertain classifications to a realistic minimum.