Application of Multiple Machine Learning Method (Hybrid Model) to Detect Phishing Links as Preventive Measure in Minimizing Victims of Data Theft

Abstract

The hegemony resulting from the widespread use of technology and social media cannot be stopped. On the other hand, the negative effects of an interconnected world are always present in the minds of all users. As with technology, crime evolves over time, creating new spaces that were previously uninhabited. Phishing, one of its forms, has become a terrifying spectre in the digital world because it exploits the greatest vulnerability in cyberspace: humans. This journal aims to present the results of the author's research into developing a machine learning (ML) model to assist in detecting and recognizing phishing links, while also proposing the use of ML by the National Police to assist citizens in avoiding phishing crimes. The research was conducted using the ML approach with three (three) methods, namely Support Vector Machine (SVM), Decision Tree, and Random Forest separately and their combination (hybrid model) in four (four) assessment areas, namely the level of accuracy, precision, recall, and scores f1.