Phishing Detection via Identification of Website Identity

In this paper, we propose an anti-phishing method to protect Internet users from the phishing attacks. The scope of our study is on the Internet phishing, particularly focusing on the detection of phishing website. In order to do that, our proposed method will render a screenshot of the webpage and segment the region of interest, which consists of the website logo. Next, we will utilize Google image database to identify the website identity based on the segmented website logo. During the identification process, we employ the content-based image retrieval mechanism provided in Google Image Search engine to locate the most similar logo from Google image database. The returned results will reveal the real identity of the website. With the real identity, we can differentiate a phishing website from the legitimate website by assessing the domain name of the query website. The conducted experiments show promising results and our findings prove that we can effectively detect a phishing website when we manage to determine the real identity of a website.