On the development of Naive Bayes classifier with application to email filtering

Abstract

It is to everybody's knowledge that the Internet is widely used around the world and communicate with others via email is of the utmost importance. For student, they can send the email to ask question and teachers reply the answer. For businessman, they use email to communicate with other business partners that in other places or countries and send some urgent documents through email. It is no doubt that email provides a platform to client to exchange messages and files within a short time by using computer. However, as the convenience of sending email, some corporations or people send a lot of advertising mails to people everyday or within a short period. It is very annoying when we find that the mailbox is filled up of these mail (we called it spam). Hence, a filter is required in the classification of mails. In this report, the application of Naïve Bayes classifier to email filtering is studied. It shows how to apply Naïve Bayes classifier in spam filtering. In addition, Gmail contextual gadget is used to study the effectiveness of Bayes spam filtering and to show that it can apply in the browser.