Feature Selection based Classification of Spams Using Fuzzy Support Vector Machine

Abstract

E-mail is a safe way for the online form of communication as well as for transferring data or messages through the web. A standard increase in popularity, which gives rapid growth to unsolicited emails in bulk amount Numerous methods are available to filter the data which is to detect automatically as well as eliminate such untenable messages. There aoe a large number of techniques for email spam filtering such as clustering techniques, learning-based technique, knowledge-based technique, heuristic processes and many more. The research of spam filtering by using the feature selection through relief feature selection technique and then classifies data by using a fuzzy support vector machine to deal with the factors of uncertain type. The experimental outcome illustrates that given algorithms can improve the filtering rate parameters and spam filtering speedmore efficiently.