An Adaptive and Dynamic Methodology for Detecting Phishing Email Based on Artificial Intelligence

Abstract

Phishing is an exploit method that is used to steal information from a targeted individual or an organization by impersonating legitimate sites. This paper can detect such targets and warn the users. Classified mails are used to train the classifier for classifying the email and the others are used for testing the derived classifier model. The feature vector of the training set is trained by extracting and breaking down elements from the mail. Training is done so as to obtain required classifier model which is then used to classify the mail into an phishing email or not. The main objective of this project is detecting phishing through email address separation. This phishing detection method is found to be effective against various targeted phishing attacks. The proposed system is able to check and verify the phishing emails that usually target in large numbers by impersonation of a particular organization so as to be able to gain access to their resources by flagging them as phishing mail or not.