Abstract: In this paper, we utilize a simple support vector machine to identify commercial electronic mail. The use of a personalized dictionary for model training provided a classification accuracy ...