Spam Filters


Dangers and benefits of using spam filters

Spam filters are the newest addition to the Internet security software family. Considering the fact that we often spend a few hours per month on removing unwanted emails, the growing popularity of spam filters is hardly surprising.

How do spam filters work?

Generally, all spam filters do the same essential thing: they "read" emails before you do and, based on various spam filtering methods the use, they decide whether a message is 100% spam, a potential spam e-mail or a valid message.

Most of the spam filters will send a challenging message to those senders who are suspected of being spammers. The filters will allow the mail from those address only if the sender answers appropriately to the challenging message.

The methods they employ, however, may be very different. The most popular ways of determining if a message is a spam or not are the following:

(1) Bayesian filter - based on an old, but extremely efficient mathematical formula, the program decides if the words used in the message's body are likely to be spam or not. While such filters are prone to making mistakes in the beginning, they are able to "learn" from their mistakes. A well-trained Bayesian filter is a formidable weapon against spam.

(2) Whitelists/blacklists - the most popular method of determining if a message is spam or not is checking the sender's address. If it is included in your mailing program's contacts, the message is probably a valid message. If it has been sent from an unknown address (or one that was known for sending spam before), the message is probably spam.

(3) Keyword search - another popular method of spam detection. If a short email contains the word "Viagra" five or six times, chances are that it is spam.

(4) Oddities search - a popular and useful feature of many spam filters. In addition to other methods, they check an email for oddities such as intensive use of incorrect HTML tags or suspicious typing mistakes ("0" instead of "o" or "3" instead of "E"). Most of the time, such typing mistakes are a sure sign of spam.

(5) Sender's IP check - Using the sender's IP as a means of determining whether the message is spam or not. The chances are that if you receive a message from a small country far away, it will be spam. Of course, this method is used only in conjunction with other filters - otherwise it would be impossible to conduct business abroad.

These methods are only a few examples of ways to deal with spam. There are many more in a spam filters' armory (sending challenging messages, checking the message titles or attachments and many others).

Choosing the best spam filters

There isn't really a "best" spam filter, although there are some programs that are above average (examples: Cloudmark Desktop or Spam Bully). However, if you want to get the most for your money, you should probably follow these simple rules:

*do not buy anything that relies on a single filter. It will lead to many false-positives, especially in the beginning.

*buy only those spam filters you are sure you'll be able to use and understand - some of them (G-Lock, for example) can be for advanced users only.

*don't pay attention to the advertisements, read some independent comparisons and reviews, instead. While different tests produce different results, the chances are that good software will tend to appear somewhere near the top.


All content copyright 2006-2017, Bonobo Pty Limited. All Rights Reserved.
Privacy Policy | Terms Of Service