iEntry 10th Anniversary RSS Archive

IT Management Begins With Security
SecurityProNews > News > Security News > MX Logic Adds New Classifier To Spam-detection System
Search:
[ news_security_news ]

MX Logic Adds New Classifier To Spam-detection System



SecurityProNews
Staff Writer
2004-09-01

SecurityProNews: Insider Reports Insider Reports RSS Feed


Effective Email Defense Must Be Flexible Enough to Incorporate Constantly Evolving Filtering Techniques.

MX Logic today announced the addition of a new classifier -- reputation analysis -- to its patent-pending Stacked Classification Framework spam-detection system.

"Reputation analysis is an important weapon in the anti-spam arsenal when used as a complement to traditional content-based email filtering," said Scott Chasin, CTO, MX Logic. "The beauty of MX Logic's modular approach to email-threat protection is that it allows MX Logic to take a long-term view of email threats and provides us with the flexibility to add new filters, like reputation analysis, as they are developed, ensuring the most effective and up-to-date defense against the devastating effects of unwanted and harmful email."

Reputation analysis rates the reputation of the email sender based upon the percentage of spam messages previously sent from that Internet Protocol (IP) address. The greater the percentage of spam messages sent from a given IP address in the past, the more likely recent email from that same IP address is spam. The new spam classifier is powered by data collected by the MX Logic Threat Center, which analyzes millions of email messages every day and reacts to real-time information about the reputation or source of an email.

MX Logic's Stacked Classification Framework is a multi-layered system that combines the industry's most powerful spam-fighting techniques to successfully block 98 percent of spam and ensure very low levels of legitimate email marked as spam. Working within the Stacked Classification Framework, the different spam filters separately assess and "vote" on the probability that a specific email is spam.

In addition to reputation analysis, MX Logic also added URL filtering to its Stacked Classification Framework. URL filtering increases spam detection accuracy by comparing embedded links found in email messages with URLs associated with identified spam. Along with reputation analysis and URL filtering, the other classifiers in MX Logic's Stacked Classification Framework include statistical filtering, industry heuristics, proprietary heuristics and reputation-based real-time blackhole list (RBL) filtering.

"Email defense technology is evolving very rapidly. No one technology is a panacea -- they all have their strengths and their limitations," Chasin said. "Buyers should be wary of any one filtering technology being touted over others and should look for anti-spam solutions that offer diversified multiple layers of defense that can constantly incorporate the latest and greatest filtering techniques."

Chasin added that reputation analysis is likely to be refined even more as new technologies that provide more than just content-based information are deployed.

"Emerging email defense technology like authentication and third-party accreditation will allow reputation analysis to evolve further, so that the reputation scores take into account more than just content and include information on the legitimacy of the sender of a message," Chasin said. "Any effective email defense solution must be flexible enough to incorporate these emerging filtering techniques."

Email authentication and third-party accreditation allow filters to quickly recognize legitimate, authorized senders of email. Email authentication helps verify the origins of email, making it more difficult for spammers to stay in business. Specifically, authentication protocols, including Sender ID, allow for the publication of Mail Transfer Agent (MTA) authorization records in the Domain Name System (DNS). Third-party accreditation tools allow authorized email senders to include information in their email header that signals to filters that the messages are not spam.

Robust spam, virus and worm detection effectiveness is critical as the volume and costs associated with email threats continue to rise. Ferris Research estimates that in 2003, the total cost of spam to U.S. businesses in lost productivity was $10 billion. In the month of July, the MX Logic Threat Center reported that 84 percent of all email traffic filtered by the company was spam.



About the Author:
SecurityProNews is a daily online and email publication focusing on internet security issues.

More news_security_news Articles

SecurityProNews: Insider Reports Insider Reports RSS Feed


Get Your Site Submitted for Free in the World's Largest B2B Directory!

Email Address:
* URL:
*
*Indicates Mandatory Field

Terms & Conditions

iEntry Featured Services: Jayde Member Services | Forums | Freeware | Advertise with Us

Virus Warnings

Subscribe to
SecurityProNews FREE!



[ more newsletters ]

article resources
Search Articles:
[advanced search]

WebProWorld.com
Get in-touch with industry experts and leaders
Post your site for review by expert and peers
Ask Security, IT, Development and Design questions

Free Membership: Join Now!

Visit WebProWorld.com

Titan Quest Forum
The #1 Titan Quest forum
Halo 3 Forum
The best Halo, Halo 2, Halo 3 forum
Nintendo Wii
Nintendo Wii news and views
Mac Software
The best in OS X freeware
Graphics Forum
Your source for graphic tutorials
SecurityProNews.com | Breaking eBusiness News Get Your IT Questions Answered - Click Here SecurityProNews News Feeds