Automated Filtering Solutions

Facts: Many of today’s anti-phishing filters are spam filters that have been retrofitted to also catch some phishing emails. Some of these solutions rely primarily on manually maintained lists of phishing emails and phishing websites. As a result, these solutions fail to catch a significant number phishing attacks, including many spear phishing attacks. They also tend to always be one step behind given the time it takes to update the so-called blacklists many of them rely on. Finallly, many of these filters have also been shown to have unacceptably high false positive rates, making them highly ineffective in practice.

 

Wombat’s PhishPatrol™ is a phishing email filter that uses advanced machine learning techniques to catch phishing emails. In contrast to solutions that rely solely on blacklists and email signatures, PhishPatrol's advanced machine learning technology enables it to build heuristics that work from the very start of new attacks. Extensive evaluation of PhishPatrol™ has shown that it consistently catches more phish than the best email spam filters and has far fewer false positives. PhishPatrol™ has been designed to easily integrate into a variety of environments and complement a company’s existing filters.

Scientific Publications

PhishPatrol™ is an enhanced version of the Pilfer anti-phishing technology originally developed at Carnegie Mellon University:

Learning to Detect Phishing Emails

I. Fette, N. Sadeh, and A. Tomasic. In Proceedings of the 16th International conference on World Wide Web, Banff, Alberta, Canada, May 8-12, 2007.