Phishing Detection Method Based on Borderline-Smote Deep Belief Network

With the rapid development of Internet, phishing and other frauds are becoming more and more serious. Criminals posing as banks, electricity providers, social networking sites to send fraudulent information to induce users to log on, steal user information, so that the vast numbers of users and financial institutions suffered property and economic losses. How to accurately and effectively identify phishing related Internet risks has been a major concern of the Internet. This paper analyzes the development history of phishing prevention and control, and presents a Borderline-Smote (Synthetic Minority Over-sampling Technique) DBN (Deeping Belief Network) method to detect phishing. The method uses deep learning phishing detection method based on web documents content and other features to improve 1% on the recognition accuracy. Furthermore the paper uses Borderline-Smote to solve the imbalanced data problem in the training of phishing detection, and further improve 2% on the F-value and recall rate.