Has wading through your inbox become as tedious as waiting in line at the DMV? Do you find yourself wondering what type of person would buy Viagra or trade stocks from random emails?

It’s hard to fathom that SPAM is the problem that it is, but disbelief won’t clear your inbox from this pesky problem.

The good news is that a solution is possible. There are basically three types of SPAM filtering technology:

• RBLs (real-time black lists)
• Heuristics (computer instructions) 
• Firgerprinting (third party approach)

First, a RBL is simply a list of domains (web-sites) that are known for sending spam.  The idea is that if you are caught sending SPAM you get put on the RBL and are thus blocked from anyone using the list. Unfortunately, spammers have figured out clever ways to outsmart the filters, such as constantly changing their domain names. In some cases, they will actually breach the security of your domain and hijack your own mail server to send out their SPAM.

Heuristics is a computer science term. In the context of blocking SPAM, it refers to an algorithm, or complex set of computer instructions, that decides what is legitimate e-mail and what is SPAM. For example, an algorithm might flag words, phrases, or character strings like “Mortgage, Viagra, $$$$, HOT STOCK, etc.”  There are many more colorful examples, but I think you get the point. The problem with this approach is that algorithms aren’t perfect and oftentimes good e-mail gets categorized as SPAM. We call this a “false positive.” When that happens, you either lose out on important information or have to dig through your spam folder looking for miscategorized messages.

The third approach, and in my opinion the best, is the use of “fingerprinting.” With this approach a third party sets itself up to receive as much SPAM as possible. The third party receives, identifies, and takes a “fingerprint” of hundreds of thousands of SPAM messages. This works because SPAM messages are not unique. The sender crafts a message that they will send to millions of recipients, over and over and over again. Once the message has been identified as SPAM, it can be filtered by everyone subscribing to the fingerprinting service. The key advantage of this methodology is that no unique message will ever get blocked - thus eliminating the problem of false positives. My own company has had great success with a product called Cloudmark

Tags: Email, RBL, SPAM, SPAM filtering technology

This entry was posted on Wednesday, June 11th, 2008 at 12:26 pm and is filed under Email.
You can follow any responses to this entry through the RSS 2.0 feed.
You can leave a response, or trackback from your own site.