Bayesian Spam Program

Written by Amy Hall
Bookmark and Share

The Bayesian spam program is based on mathematical algorithms and specific probabilities that can filter out spam or junk email. By incorporating several restraints and classifying text, the email filter is able to greatly reduce the amount of junk email that makes it to your inbox. Using the Bayesian method is relatively new and its uses continue to be explored.

Nothing is Fool Proof

Even though the Bayesian spam program has been through many tests and continues to be improved, there is no guarantee that junk email will not make its way past the filter. One trial performed by noted developer Paul Graham had a success rate of 99.75%, only allowing 4 spam messages of 1,750 to his inbox. While the results are very impressive, developers need to continue to make improvements because spam senders are becoming smarter.

Mr. Graham was able to identify the reasons for the messages making it to his inbox in order to modify his Bayesian spam filter. While three of the messages could be identified in the future as junk mail, one message led him to believe that he had found the spam message of the future. The text of the message was generic and neutral followed by a website address to follow. The address turned out to be an ad for an adult website.

The Future For Bayesian Spam Program

Being able to filter out the majority of junk email is not always sufficient as solicitors will continue to take advantage of cracks in the system. In order to make a better filter it is important to understand how spam messages made it through the filters and figure out what could have been done to prevent it. This type of work is usually not for the technically challenged but consumers can take advantage of the numerous email filtering programs available.


Bookmark and Share