VARIOUS SPAMS AND CLASSIFICATION ALGORITHMS FOR DETECTION OF SPAM EMAIL THROUGH COMPARING J48, SVM AND NAIVE BAYES CLASSIFIERS USING WEKA TOOL

Abstract

Electronic Mail (E-mail) is playing most important and significant role taken in the world of information communication for users. Nowadays, Email is most common and effective mode of communication technology for communicate and sharing the information to both end users. The rapid increase of email users there will be increase of volume of spam emails too from the past few decades. Emails are categorized in to ham and spam emails. This paper illustrates on different existing email spam filter system regarding Machine Learning Technique (MLT) such as Naive Bayes, SVM, J48 and present the classifications, evaluation and comparison on different email spam filtering algorithms using WEKA Software and performs various parameters like finding Accuracy, Recall, Precision, Measures and False Position Rate etc. The final output result should be ‘1’ if it is finally spam present, otherwise, it should be ‘0’ for non-spam. In this analysis the Final out presents that J48 classifier is best and efficient algorithm for spam or not spam emails among other algorithms.