Abstract: Phishing is a fraud attempt in which a scammer acts as a trusted person or reality to gain sensitive information from an internet user. In this Methodical Literature check (SLR), different phishing discovery approaches, videlicet Lists Grounded, Visual Similarity, Heuristic, Machine Learning, and Deep Learning grounded ways, are studied and compared. For this purpose, several algorithms, data sets, and ways for phishing website discovery are revealed with the proposed exploration questions. A methodical Literature check was conducted on 80 scientific papers published in the last five times in exploration journals, conferences, leading shops, the thesis of experimenters, book chapters, and from high- rank websites. The work carried out in this study is an update in the former methodical literature checks with further focus on the rearmost trends in phishing discovery ways. This study enhances compendiums’ understanding of different types of phishing website discovery ways, the data sets used, and the relative performance of algorithms used. Machine literacy ways have been applied the most, i.e., 57 as per studies, according to the SLR. In addition, the check revealed that while gathering the data sets, explorationers primarily penetrated two sources 53 studies penetrated the PhishTank website (53 for the phishing data set) and 29 studies used Alexa’s website for downloading licit data sets. Also, as per the literature check, utmost studies used Machine literacy ways; 31 used Random Forest Classifier, which, as per different studies, achieved the loftiest Accuracy, 99.98, for detecting phishing websites.

Keywords: Phishing, URL, Hyperlinks, Machine Learning, Random Forest, K-means, SVM.

| DOI: 10.17148/IARJSET.2024.11711