Detecting phishing pages using the relief feature selection and multiple classifiers

Abstract

Website phishing is a deception in e-commerce, which attempts to steal user confidential information using similar websites. The classification technique is one of the common ways to detect phishing websites. According to high-volume main data, attribute reduction algorithms play an essential role. This paper presents an appropriate model based on the relief algorithm to reduce dimension. Moreover, the proposed approach uses multiple-classifiers to increase accuracy. The evaluated results show higher accuracy and superiority than conventional methods.