Phishing Detection in Websites using Parse Tree Validation

Phishing is a technique of tricking people into giving sensitive information like usernames and passwords, credit card details, sensitive bank information, etc., by way of email spoofing, instant messaging, or using fake web sites whose look and feel gives the appearance of a legitimate website. In this work, a technique named parse tree validation is proposed to determine whether a webpage is legitimate or phishing. It is a novel approach to detect the phishing web sites by intercepting all the hyperlinks of a current page through Google API, and constructing a parse tree with the intercepted hyperlinks. This technique is implemented and tested with 1000 phishing pages and 1000 legitimate pages. The false negative rate achieved was 7.3% and the false positive rate achieved was 5.2%.