University of Osnabrueck
Undesired e-mail (spam) becomes a big problem nowadays not only for users, but also for Internet providers. One of the main obstacles for elimination of this problem is a complicated security issues. In particular, the very low rate of falsely detected e-mails in practice. The authors tried to eliminate this problem by a sufficient similarity check of checked e-mails with e-mails marked as unrequested or spam. This paper uses Bayesian algorithm with many variations. For comparison purposes, they used a normalized compression distance which helped the user reduce the rate of false detection of individual mails.