Spam Mail Classification Using Combined Approach of Bayesian and Neural Network

Abstract

Unsolicited commercial e-mail (spam) has shocked economies world over and is threatening the productivity. In this paper, an attempt has been made to classify email spam by combining Bayesian network and neural network classification approach. The header information like sender details and origin IP etc. was analyzed by centered Bayesian network, whereas the content and subject of the email were separately analyzed to classify the e-mail by neural network as a classifier trained by genetic algorithm (GA).