EPG VPG Support Services Technology Testimonials Success Stories Channel Partners Technology Partners Become a Partner Partner Portal Press Releases Events In the News Associations Press Room Overview Global Citizenship Leadership Careers Contact Us

Abaca's® Filtering Technology

Breakthrough technology redefines spam protection.

Receiver Reputation

The unparalleled success of Abaca’s Email Protection Gateway lies in its unique approach to detecting spam. The vast majority of today’s anti-spam providers attempt to detect spam by analyzing either characteristics of the sender or the e-mail content or both. Abaca’s revolutionary solution is to analyze the distribution of spam versus legitimate email for the message recipients and calculate a receiver reputation for each from that distribution.

The concept of receiver reputation is based on the fact that different people receive different amounts of spam and legitimate email. When analyzing a message, each receiver’s percentage of spam versus legitimate email (his or her reputation) is an estimate of whether the message is spam or legitimate. Essentially, if the message is sent to users who typically receive a high percentage of spam, the message is more likely to be spam. However, if the message is sent to users who typically receive a low percentage of spam, the message is more likely to be legitimate. Aggregating the reputations of all recipients of a particular message, therefore, is equivalent to combining those users’ rating power to estimate the legitimacy of the sender and the message. In a receiver reputation system, the key determinant of whether a message is spam or legitimate is not the identity of the sender or the content of the email, but the reputations of the email recipients, individually and collectively.

All Users Participate in Identifying Spam

Suppose we separate the world into five groups based on the amount of spam they receive on a daily basis. People in Group 1 receive, on average, 90% spam. Group 2 receives 70% spam, Group 3 receives 50% spam, Group 4 receives 30% spam, and Group 5 receives 10% spam. The two graphs below demonstrate how a receiver reputation system works when legitimate emails and spam emails are sent to email users in these five groups.

The first graph shows the distribution of 25 legitimate emails sent from a given IP address to users comprised of members of the five groups described above. The positive slope of the line connecting the blue shaded bars indicates a high likelihood that the message is legitimate. The second graph shows the same distribution for 25 spam messages sent from a given IP address to members of the same five groups. The negative slope of the line connecting the red shaded bars in the second graph indicates a high likelihood of a spam email message.

Receiver Reputation Mathematically Guarantees Superior Results

It is a mathematical certainty that for any sufficiently large mailing (ten messages or more), the message distribution will appear as in one of the chart above. Spammers send billions of messages each day, meaning they cannot escape this mathematical certainty and will be detected. Thus, a receiver reputation system typically identifies a spam in fewer than ten messages, even if the message content is unique and defeated every known checksum. There is simply no way for a spammer to escape detection in a receiver reputation system. Messages are rated by WHO the spammer sends messages TO, rather than where the message is FROM or what it CONTAINS. The receiver reputation system is infallible because spammers must send to people who, in aggregate, get more spam than the average email user. This is a mathematically guaranteed fact that a spammer cannot defeat.

ReceiverNet: A Breakthrough in Spam Prevention

Abaca has redefined spam prevention. The core engine behind Abaca's technology is ReceiverNet, a patent-pending, receiver reputation-based approach to detect spam. The technique is new, unique and revolutionary.

ReceiverNet is based on a sophisticated mathematical formula that uses receiver reputations to precisely differentiate spam from legitimate messages. A message is considered more likely to be legitimate if it is sent to recipients that typically receive a low percentage of spam. Conversely, a message is considered more likely to be spam when sent to recipients that typically receive a high percentage of spam.

It is not necessary to manage complicated rules, whitelists, or blacklists. Because message ratings are based on each user's overall legitimate/spam ratio (as measured by the system), users do not need to help the system identify spam other than to express personal preferences, if they so desire. The system learns and becomes more accurate on its own by tracking the legitimate/spam statistics for each protected user. Spam detection becomes more accurate as more users are added to the system.

If a message is sent to 100,000 protected users, the system has the rating power of 100,000 receiver reputations to rate the sender and the message. In practice, a spam attack is typically blocked before a protected user receives the first email. By the time a spammer has sent three messages, there is a 99.9 percent certainty that the spam message will be blocked.

ReceiverNet Advantage
Key benefits include:

• Unprecedented accuracy. Over 99 percent spam blocking means fewer than one mistake in every 100 messages processed. That’s 10 to 100 times fewer mistakes than any other available systems.
  
  
• Instant reaction time. Because ratings are based on the most recent 25 emails for each sender, the system reacts instantly to spam attacks, usually within just a few messages.
  
  
• Spammer proof. There are no complicated rules or subjective blacklists that need to be managed. Spam or legitimate email is based on a mathematical relationship that a spammer, who must send in volume, cannot avoid no matter how hard he tries.
  
  
• Language independent. Focus is on the sending IP-address and who the receivers are, not the message content.
  
  
• Highly decisive. Virtually all messages are rated as a strong ham or strong spam. Fewer than 4 percent of incoming emails fall into a grey area where the system has not yet seen enough samples to make a definitive rating. These are called out for the user to decide.
  
  
• Immune to human rating errors. Humans are notoriously bad at rating emails. Our system doesn’t rely on human opinion, but only uses each receiver’s aggregate spam percentage in our calculations.
  
  
• No training required. There is no “ramp up” period. The system determines the spam percentage of each new user from the ratings established by all existing users.