Detecting the Phishing Website with the Highest Accuracy

Abstract

Phishing attacks are increasing and it becomes necessary to use appropriate response methods and to respond effectively to phishing attacks. This paper aims to uncover phishing attack sites by analyzing a three-module set to prevent damage and reconsider the awareness of phishing attacks. Based on the analyzed content, a countermeasure was proposed for each type of phishing attack by using website features. These features will be classified in order to determine the effectiveness of the countermeasure. Finally, the proposed method enhanced the site security as anti-phishing technology. The phishing detection used three classification algorithms, which are the decision tree; the supporting vector machine and the random forest were combined into one system that was proposed in this paper for the purpose of obtaining the highest accuracy in detecting phishing sites. The results of the proposed algorithm showed 98.52% higher accuracy than others.