We are all familiar with word Spam. This is something we encounter almost on a daily basis in our mail inboxes. They waste our precious time and also contribute in spreading malwares. It is expected that in the year 2008, as many as fourty trillion spam messages will be sent across the globe. This is significantly higher than the figure in 2006, when it was 18 trillion messages.



All of us are aware that email filtering software separates good or true email from the rest or spams. But these software or other appliances employed for filtering can go wrong. While a true mail may be blocked, a spam mail can easily pass through the filter to reach your inbox. If a spam mail is successfully filtered, it is generally placed in a separate folder. These folders can be searched later by the user. According to few sources, both individual users and organizations stand to lose not only their energy and time in deleting these spam mails but also have to shell out $0.4 per deleted message. This figure has been calculated after a lot of calculations and predictions by analysts.

Popular filtering techniques

A number of spam filtering mechanisms have been employed over the years to counter it effectively. Occasionally, more than one techniques are also employed. Some of these include:

1] Tarpitting:- These are services on the server which takes a lot of time establishing incoming mails. This technique ensures spam mails will take a lot of time to reach a recipient, thus discouraging the spammer to carry out his task. However, the disadvantage is that the same procedure hampers the true mails as well.

2] Graylisting:- In this case, the recipients mail system temporarily rejects all incoming mails. Reciprocal mails are sent as reply with the content being quoted as temporary failure. This technique is employed with the assumption that a real mailer will respond back by sending the mail once again while the spammer loses his patience and will give up on sending further spams.

3] Challenge response:- This technique employs CAPTCHA method to its routine. Mails from unrecongnizable IP addresses are asked to enter the sent CAPTCHA. Humans are able to detect these patterns while the computer fails to do so.

4] Bayesian filters:- This technique analyses the ratio of "good" and "bad" words and assigns a score to each one of them. Based on these scores, spams are filtered out.