Bayesian Spam Filter

Written by Jessica Duquette
Bookmark and Share

A Bayesian spam filter assigns a probability number to each word within an email. Words that are highly likely to be spam increase the probability of the message being blocked by the filter, whereas words that are most likely contained in a legitimate email will decrease the chances of the message getting blocked. This is the foundation for the content-based filter known as Bayesian.

A Bayesian Spam Filter Learns From You

While everyone's email is different, the most successful filter will adjust to the email habits of the end user. If you typically receive messages that contain the word "sandbox" then you're filter will determine that most emails containing that word should not be filtered or blocked. This form of intelligent Bayesian spam filter only increases its accuracy and productivity the longer you use it.

An easier way to ensure you're getting all of the email from your regular contacts is to add them to a white list. The opposite of black list, the white list contains the names and addresses of the people whom you normally exchange correspondence. By adding their address to a white list you ensure that all messages from these people are passed through the filter.

Black & White Lists

Your black list, or block list can also filter all email addresses you wish blocked. Many email clients such as hotmail will also block any emails coming from a similar domain name. Unfortunately advertisers are getting more creative and have now figured out a way to email each message from a different address, making a white and black list obsolete without a spam filter.

Bookmark and Share