URL Based Phishing Website Detection using Machine Learning Model

• Python
• Machine Learning

Phishing is a cyber security attack that is used to trick victim users to provide sensitive information or deploy malicious software on their infrastructure. Depending on the target system and users, these attacks can inflict severe negative impacts on the system. Therefore, researchers have been working on developing phishing detection and prevention techniques to thwart these attacks. In this paper, we present an efficient phishing websites detection system that analyzes the phishing websites URL addresses to learn data patterns that can identify authentic and phishing websites. Our system employs machine learning techniques such as neural networks and decision trees to learn data patterns in websites URLs. We evaluate our system on a recent phishing websites dataset using classification accuracy as a performance indicator. Our best result shows that decision trees models provide 97.40% classification accuracy on the almost balanced-class dataset. To overcome the drawbacks of a blacklist and heuristics-based technique, several security researchers currently targeted machine learning techniques. Machine learning technology consists of the many algorithms that need past information to create a call or prediction on future information The algorithmic rule can examine multiply prohibited and real URLs and their attributes to properly observe phishing websites, together with zero-hour phishing websites, victimization this approach.