A Hybrid Bayesian Approach with ABC to Recognition of Email SPAM

This paper presents a hybrid Bayesian algorithm using Swarm intelligence to recognize and block SPAM Emails. Electronic mail is widely used for personal and business communication. Due to low cost communication by using emails for sender, several people and companies use it to quickly distribute unsolicited bulk messages, also called spam, to a large number of recipients. Due to unnecessary traffic and due to security threats, Spam has become a major threat for business users, network administrators and even ordinary users. In addition to regulations, several technical solutions have been proposed and deployed to block this problem. Among all content based SPAM recognition is best suited. By including Bayesian rule (Baye’s Theorem) in content scanning, over all throughputs for recognition of SPAM mails can be increased. But In this approach limitation is also there due to static values of probabilities for each token. So automated training is required for filter. A strong automated trained filter can also maintain by including Nature based optimization techniques like ABC (Artificial Bee Colony Optimization), SMO (Spider Monkey Optimization). In which best tokens can be classified to recognize the SPAM. KeywordsSPAM, SMO, ABC, PSO, Email.