An entice resistant automatic phishing detection

Phishing is turning into a hotbed for vast fraudulency over the Internet; therefore it's one of the most challenges toward Internet security. Utilizing a centralized list of Website is a common solution; as the most of the browsers and commercial anti-phishing products utilize it. Nevertheless, this solution is helpless against zero-day phishing attacks. So, many researches study and suggest methods based on machine learning techniques. Most of the features involved in these methods can be easily enticed. This paper introduces a novel method with high precision and also resistant to enticement. This method was tested against common legitimate and phishing websites, and produced high detection accuracy.