MailFrontier Presenting at 2005 Spam Conference
MailFrontier Selected to Present Two Research Papers at Third Annual Academic Conference

Palo Alto, Calif. –January 20, 2005 – MailFrontier™, a leader in email security, will participate in the 2005 Spam Conference being held on January 21st at Massachusetts Institute of Technology (MIT). As the only email security vendor invited to participate, MailFrontier will present research on spam blocking and phishing detection technologies. The Spam Conference gathers leading researchers from academia and industry to share ideas and techniques for protecting email systems and users.

WHO/WHAT:	Jonathan Oliver, Director of Research, MailFrontier “Using Lexigraphical Distancing to Block Spam” Presentation> Andrew Klein, Product Manager, MailFrontier “Bayesian Spam Classification Applied to Phishing Fraud” Presentation>
WHEN:	Friday, January 21, 9:00 am – 5:00 pm ET
WHERE:	2005 Spam Conference MIT, Room 26-100 Cambridge, Mass. http://spamconference.org/

Oliver will present a paper entitled “Using Lexigraphical Distancing to Block Spam.” Spammers are aware that content-based email filters analyze text to block spam, so they often use mutated or morphed content. This morphed content is understandable by human eyes, but can trick anti-spam blocking systems into letting the email through. Oliver will discuss an adversarial-parsing technique called lexigraphical distancing which can correctly identify the nature of the mutated or morphed expressions. Research results demonstrate how this approach significantly increases the effectiveness of spam filtering.

Following Oliver’s presentation, Klein will present “Bayesian Spam Classification Applied to Phishing Fraud,” as a method to increase the accuracy of phishing filters. While spam is a marketing email designed to call attention to itself, phishing emails are specifically designed to resemble good email making them difficult to catch using spam filtering technology. MailFrontier pioneered the tailoring of Bayesian classification, a common spam filtering method, to identify that an email is a phishing fraud. This presentation will discuss how Bayesian email filters can be modified to correctly identify phishing emails.

About MailFrontier
MailFrontier is a leading provider of email security solutions that stop unwanted email by protecting enterprises from the growing number of dangerous and costly email threats while giving them the ability to anticipate and stop new threats as they emerge. MailFrontier offers a comprehensive defense against all known email threats, including spam, viruses, phishing, blended threats and other attacks, such as directory harvest attacks. The company has more than 850 customers, including Peet’s Coffee & Tea, Pier 1 Imports, PNM Resources, San Francisco Giants and Wyndham International. Founded in February 2002, MailFrontier is headquartered in Palo Alto, California. For additional information, go to www.mailfrontier.com.

###

MailFrontier is a trademark of MailFrontier, Inc. All other company and product names are trademarks or registered trademarks of their respective holders.