Bayesian Spam Filter Based on Distributed Architecture

Abstract

The flood of spam promotes the development of anti-spam technology. In this paper, we bring forward the Bayesian filter technology based on the distributed architecture, which can realize the sharing of the Bayesian learning outcomes among servers within the system, so as to increase the accuracy of spam recognition. We, in the paper, discuss the sharing model of information with spam features under the distributed architecture and the spam identification process; analyze the Bayes algorithm and carry out the relevant improvements; design the Bayes Filter based on distributed architecture on the above basis and verify the effect of the filter by experiments.