Spam Filtering at the Client E-mail Level

Abstract

Spam has now become a significant security issue and a massive drain on financialresources. In this paper, a method for filtering the spam at the client level is presented.The proposed filter combines more than one filtering mechanism that would make thefilter more efficient, faster and low false positive. The main mechanism implemented isthe Bayesian filter combined with a blacklist and whitelist. The header of the incominge-mail will be tested against the whitelist to determine whether the e-mail is legitimate ornot. Also it will be tested against the blacklist to determine whether the e-mail is spam ornot. In case of no matched results the e-mail will be checked by the Bayesian filter. Theresults of this check will be then used to update the whitelist and blacklist