Detection of Phreaking Website Using Various Algorithms

Abstract

A big concern to the Internet nowadays is phishing, a crime that involves exploiting technological tools to steal sensitive consumer data. Phishing losses are also rising quickly. The importance of feature engineering in solutions for detection of phishing websites, however the precision of detection is crucial and it depends on the features you know already. Additionally, although features retrieved from multiple dimensions are more thorough, extracting these characteristics has the downside of taking a long time. To address these, we proposed a new approach in which dataset contains millions of URLs by this approach we can identify the URL which is attacked by the phisher. To determine whether the URL has been targeted by the phisher, some of the Convolutional Neural Network algorithms like CNN-LSTM, CNN BI-LSTM, Logistic Regression, and XG Boost are utilized and resulting in the correctness of the graph between the two machine learning methods by using trained dataset and more likely to produce sensitivity, specificity, precision, recall, and f1-score along with accuracy graph, confusion matrices and also along with ROC-AUC curves.