Myanmar Spam Filtering based on Naïve Bayesian Learning Algorithm (MSFNBLA)

Abstract

Spam is one of the main problems in e-mails communication. Naïve Bayesian (NB) learning is useful algorithm for constructing Myanmar spam corpus (SCorpus) from pre-defined spam and ham e-mails. SCorpus is built on the assumption that the characteristics of e-mails in the training dataset. Content-based analysis is particularly effective in filtering spam. NB plays a critical role in probabilistic learning and calculates the probability of an e-mail being spam based on its contents. The motivation for this paper is to find a solution for the Internet users in Myanmar e-mails received every day in their mailboxes. There is no standard implementation for treatment of Myanmar emails. So, a classification filter for the e-mails should be proposed with SCorpus. NB approach is being popular for learning corpus. To filter the spam e-mail, paper MSFNBLA is applied for classifying the incoming e-mail is spam or ham.