Anti-spam experts at the world's number one Web and e-mail filtering company, SurfControl, released the top five ways proven to identify and stop spam in the enterprise.
The methods, state-of-the-art technologies available to today's IT executives, are identified in a new white paper, "Major Techniques for Classifying Spam."
Written by a team of e-mail filtering experts at SurfControl, the white paper comes as more and more people agree that it will take a combination of advanced technology and a new legal infrastructure to effectively fight spam.
Paris Trudeau, a White Paper author and SurfControl E-mail Filter Product Marketing Manager who regularly advises enterprises on ways to combat spam, says that increasingly effective software tools are capable of correctly identifying and filtering a large percentage of spam e-mails.
"The key," Trudeau says, "is to find ways to maximise the accuracy of identifying spam with the least amount of human intervention, which can be enormously costly. To do this, organisations need to use multiple spam detection and classification technologies that analyse and filter actual e-mail content automatically and in real time to stop spam at the network edge."
Customisation is also important, she added. "One company's spam is another's legitimate business e-mail."
Top Five Ways Proven to Classify and Stop Spam:
1. Fingerprint Database Analysis
2. Lexical Analysis
3. Artificial Intelligence
4. Statistical Analysis
5. Heuristics
Fingerprint Database Analysis for spam is similar to fingerprint-based identification common in anti-virus software, and is very effective in identifying known spam strains and is unlikely to identify an innocuous message as spam. Lexical Analysis examines words and phrases within message content in the context of the whole e-mail, which is necessary, the authors note, "because spam messages are constantly mutating to avoid detection." Artificial Intelligence using neural networks that can be trained to learn what an organisation defines as spam.
Statistical Analysis, a method similar to AI, can be trained to weight the overall probability that a message is spam. Heuristics, a term with Greek origins that means "to find," is a framework that combines the results of spam identification tests, determines an overall score of a message's content and ultimately identifies a message as spam or not.

Comments
Post new comment