28 Jul 2014
The Evolution of Anti-Spam Filters Over Time
28 Jul 2014
These days, everyone knows what spam is and nearly everyone has an email address that receives spam daily. But this wasn’t always the case.
The first email spam in history was sent by Gary Thuerk to 393 people in 1978. At the time, it was necessary to send the message individually to each person. Thus, no filter was implemented to protect inboxes from receiving spam.
Until the first half of the 1990s, few major spam campaigns were reported. Before this period, the community attempted to self-regulate.
In 1993, the word …spam“ appeared for the first time to describe unsolicited messages. The first major incident took place one year later with a massive spam campaign on Usenet (a network of forums).
Between 1996 and 1999, the number of email users grew from 25 to 400 million. This growth gave birth to an industry dedicated entirely to spam.
The blacklisting technique consisted of blocking IP addresses (unique identification numbers on internet) of servers sending spam. These lists were used by the ISP and email providers to filter emails before they could arrive in user inboxes.
Unfortunately, filtering emails based on IP addresses was not enough! In addition to blacklists, which are still used today, filters based on the content of emails sent came to be. The most well known type of content filter isSpamAssassin, which was created in 2001.
This type of filter gives a score based on numerous criteria by analyzing the content and the header of the email. These scores are then culminated to give the email an overall score, which determines if it should be considered spam or not.
In 2003, the number of email users reached over 600 million. The use of blacklists and content filters were not sufficient in stopping the growth of spam. Spammers were evolving everyday to counter new anti-spam technology, and the American legislation against spam (the CAN-SPAM Act) established in 2003 was not going to be the solution.
For these reasons, two new technologies were created to allow for better identification of email senders. These were SPF (Sender Policy Framework) in 2005 and DKIM (DomainKeys Identified Email) in 2007.
The latest evolution in anti-spam filters is the use of reputation scores for IP addresses and domain names. This score changes constantly as a result of campaigns sent and is calculated by ISP and email services providers, as well as by specialized companies such as ReturnPath and Cisco.
Reputations in email marketing are not a recent concept, but the technique is evolving constantly. More and more, it takes the behavior of the email user into account (deletion before reading, opens, clicks, marking as spam…) and considers the reputations of domain names more than ever.
For an exhaustive description of different anti-spam filtering techniques, don’t hesitate to refer to this Wikipedia article: http://en.wikipedia.org/wiki/Anti-spam_techniques_(e-mail)
Photo credit: Mutt – GNU General Public License