Phishing Website Detection Using the Decision Tree Algorithm Method

Abstract

Along with the increasing number of internet users and the rapid development of technology, cyber security threats are becoming more complex, including phishing threats that often cause major losses such as loss of individual or corporate privacy. This study aims to identify phishing websites effectively through the application of machine learning algorithms. The dataset used in this study comes from the UCI learning repository developed by the University of Huddersfield. The research methodology includes the stages of problem identification, Cart algorithm collection, validation, and model evaluation. With this method, the study found that the CART algorithm was able to achieve an accuracy level of 90.5% in detecting phishing sites. These results show cyber security, especially in protecting users from phishing threats, this study is expected to contribute to improving data protection and privacy of internet users, as well as encouraging the application of machine learning technology in a more adaptive cyber security system.