Overview of Modern Spam Blockers

Modern spam blockers use various technologies to stop more spam than earlier spam fighting packages.

The first spam blockers typically relied on detecting certain text strings that commonly occurred in unwanted email. This approach is referred to as simple text filtering. While at first, it successfully detected a large portion of junk messages; as spammers became more aware of efforts to fight spam, they adapted and it became less effective.

The spam blockers in turn were adapted to account for the spammer's vigilance. In the late 90's and early 2000's new technologies such as Bayesian filtering and DNS blacklisting became common place. Bayesian filtering integrated machine learning concepts, while blacklisting used databases of known spammers to help block junk e-mail. They allowed spam blocking software to catch more spam while keeping false-positives low.

In 2002, SpamButcher introduced the first commercial anti-spam software using fuzzy logic. Fuzzy logic offers many of the advantages of Bayesian filtering without the need for training.

Recently, open-source groups along with Microsoft have introduced various standards to help authenticate message senders. While identifying the sender doesn't inherently eliminate or block unwanted e-mail, it can help control certain kinds of junk messages utilized to commit identity theft.

Back