E-mail classifications using Support Vector machine

Abstract

- Study on text categorization field contains classification process of text documents into a fixed number of pre-defined categories by user. The paper studies on the classification of email messages based on their categories using Support Vector Machine (SVM). Among classification processes involved are reading input email data from its subject and body, feature extraction process, feature selection process, and feature classification process. Feature extraction process involves word stopping and stemming methods that can reduce the number of dimension of features. We have used the term frequency inverse document frequency (tfidf) for the features selection method. The effectiveness of classification results has been measured using precision and recall criteria. From experiments, the results have shown a higher accuracy in classifying email dataset using the SVM.