A Literature Survey on Website Phishing Detection Using Data Science Models and Leveraging Various Datasets

In the modern era of computing, driven by technological advancements, data and its attributes are fueling significant shifts in technology, particularly in operations related to data analysis, management, security, and maintenance. To protect crucial data and identify and understand security incidents, it is essential to extract patterns and insights from cybersecurity data. This literature survey highlights the growing body of research in phishing detection, emphasizes the dominance of threat detection techniques, and identifies specific datasets and algorithms that have shown promising results with high accuracy. Different phishing detection approaches, including Machine Learning, Generative models, hybrid models, List-Based Models, Visual Similarity Models, Heuristic Models, and Deep Learning-based techniques, are studied and compared. Data cleaning, data balancing, feature selection, and feature extraction were carried out during the model development process to build the most sustainable model. A study and analysis were conducted on various scientific papers published in the last two years in research journals, articles, conferences, technical workshops, technology books, and high-ranking business blogs. This study enhances readers’ understanding of various phishing website detection techniques, the datasets used, and the comparative performance of the algorithms employed.