PRESENTING A NEW FEATURE SELECTION METHOD USING CFS WITH LINEAR-FORWARD-SELECTION ALGORITHM FOR E-MAIL SPAM FILTERING

The purpose of this research is presenting an machine learning approach for accuracy enhancement of detecting andfiltering automatic spam and separating them from legitimate messages. In this regard, for reducing the error rate andincreasing the efficiency, a new architecture on feature selection has been used. Features used in these systems, are the body of text messages. Proposed system of this research has used Correlation-based feature selection (CFS) withLinear-Forward-Selection algorithm. In addition, Multinomial Naïve Bayes (MNB) classifier, Discriminative Multinomial Naïve Bayes (DMNB) classifier, Support Vector Machine (SVM) classifier and Random Forest classifier are used for classification. Finally, the output results of this classifiers methods are examined and the best design is selected and it is compared with Previous similar works by considering different parameters. The accuracy of the proposed system is evaluated equal to 98%.

کلیدواژه ها:

Feature Extraction ، Feature Selection ، Classification ، Spam Filtering ، Machine Learning