Technology Overview   StarEngine   Bullet Signature Database   Technology FAQ   Technology Whitepaper (.pdf)   The Human-Touch Whitepaper (.pdf)   Products Brochure (.pdf)   Technology SDK Q. Aren’t all anti-spam solutions based on SpamAssassin? A. No – though most are. The Mail-Filters technology is 100% proprietary and is not based on any open source software. Q. How is the Mail-Filters Technology different? A. Mail-Filters has developed a unique combination of technologies to catch spam. From the beginning, we have focused on eliminating false positives, while still catching virtually all the spam. Most customers see at least 95% catch rate with no more than 1 in 1,000,000 false positive rate. The technology utilizes several techniques, but at its core is the Bullet Signature Database and the StarEngine. The Bullet Signature Database is a collection of spammer profiles created by our own human editors that identify both individual messages and future messages from the same spammer. The STAR Engine is neutralizes spammer tricks and scans messages for Bullet signatures. Q. What makes it better than other anti-spam technologies? A. 1. Human Editors 2. Non-English spam catch rate 3. Accuracy 4. Throughput speed. Human Editors allow the analysis of what is and isn’t spam to fall to our staff, instead of your administrators or users. You would expect a technology that reviews the results by human editors to be accurate, and we are. Most technologies require training on each language they work with. As a result, most anti-spam technologies don’t catch non-English spam very well. The Mail-Filters solution is a truly global solution with signatures in over 30 languages and installations in over 100 countries. The Mail-Filters technology has extremely fast throughput. Some customers have seen throughput exceed 1000 messages per second. Q. How does it work? A. The Mail-Filters technology uses a complex set of techniques to provide customers with anti-spam technology that really works. Simplified, the steps are: 1. Spam is collected in various ways. Users, automated processes, and partners provide spam in all languages 24 hours a day, 365 days a year (except leap year). 2. Human Editors review spam to determine if it really is spam and, if it is, create a profile of the spammer based on a combination of values in body, headers, envelope, and/or subject. This is not a hash or checksum database, but a unique profile that identifies the spammer, not just the individual message they’ve sent. 3. The Bullet Signature Database is updated every 1 to 15 minutes to insure new signatures catch the latest outbreaks of spam. 4. When new emails arrive, the STAR Engine detects any spammer tricks, neutralizes them, then scans the message for any Bullet Signature matches. 5. Any missed spam can be forwarded to Mail-Filters to be rated by our editors and included in the Bullet Signature Database. Q. What about new spam messages that have never been seen before – how does your technology catch it? A. One thing that’s different about Mail-Filters technology is that it tries to identify spammers, not just single messages. As you would expect, most spam is sent by experienced spammers – so we’ve got a good chance to catch their new spam messages. There’s also a good chance any missed spam messages will be collected, a new signature created, and delivered to your server, before it is delivered to your email server. Q. Most/All products claim “very low false positive rate”. I can’t afford to lose good email messages – my business depends on the good messages getting through. How do you prevent false positives? A. The Mail-Filters technology is extremely accurate because we don’t guess if a message is spam, we use real human editors in our operations centers to know whether a message is spam or not. While our true false positive rate is less than 1 in 1,000,000 messages (about 1 false positive per 3 years for the average 100 message a day person), our critical false positive rate approaches 1 false positive in 10,000,000 messages (or one critical false positive every 300 years for the average 100 message a day person). What we’re saying is that the few false positives we get are newsletters that have accepted advertising from spammers. Their ads sometimes have the same profile as their spam messages. We utilize many techniques to keep newsletters and user group messages from being detected, but are not yet perfect. Person to person correspondence is virtually never marked as spam. Q. Why do you make your technology available to other anti-spam vendors? A. We believe our technology is the best single tool for detecting spam. Others have agreed and incorporated it into their products. We realize there are many great products that might want additional anti-spam techniques and we believe they should have the best.

Copyright © 2001-2007 Mail-Filters.com, Inc. All rights reserved. Terms Of Use   |   Privacy Policy   |   Trademarks   |   Contact Us