Phishing URL Detection

Phishing URL detection involves analyzing the features of a phishing web page. This helps in detecting malicious websites that may be camouflaged with legitimate patterns. A pre-trained classifier can detect phishing sites with ease and without human intervention.

In order to achieve this feat, researchers have developed several methods. These include a machine-learning-based predictive model and a network-based inference method. They also rely on a third-party blacklist.

The most basic approach for phishing URL detection is to build a blacklist. Several third-party services are available, but the process of building a blacklist is manual and time-consuming. Another option is to implement an artificial neural network to distinguish a phishing web page from a benign one. However, a third-party blacklist has its limitations, such as the need for maintenance.

Besides, some of the heuristic features that can help in detecting a phishing page are not present in all benign sites. Thus, the best way is to perform a real-time scanning of links.

Another way to improve a phishing URL detection is to use feature engineering techniques. Feature engineering techniques are more effective because they identify patterns rather than just checking the features.

For example, a phishing URL may contain multiple dots. The character level sequences of a phishing webpage are not a coincidence. Their presence is a sign that a site contains sensitive information.

The dns_record feature is useful for phishing points, while the page_favicon feature is a nice touch. But, it does not make much sense for the result.

Leave a Reply

Your email address will not be published. Required fields are marked *