A real-time automatic detection of phishing URLs

Phishing scam, a fraudulent attempt that masquerades as a trustworthy entity to obtain users' sensitive data, has become the most dangerous form of online fraud to hit online businesses and information security. In this paper, we reveal some new aspects of the common features that appear in the phishing URLs, and introduce a statistical machine learning classifier to detect the phishing sites, which relies on these selected features. Unlike previous studies, we do not utilize an ordinary feature extraction method since some of these features need to be treated differently and some of these cannot be retrieved by the traditional way. A number of comprehensive experiments show that our proposed method achieves high accuracy over a balanced dataset and less than 1% error rates in the simulated real phishing scene with a high processing speed. And moreover, the well performance of our proposed algorithm demonstrates the new characteristics and the corresponding extraction methods are useful in the anti-phishing scenario.