Abstract: For security researchers, phishing assaults' increasing frequency continues to be a major worry. Conventional signature-based methods for phishing website detection frequently miss freshly created phishing sites. By utilizing large and varied datasets, researchers are creating machine learning-based systems that can accurately identify and categorize phishing websites in order to address this issue. Logistic Regression (LR), K-Nearest Neighbors (KNN), Support Vector Machine (SVM), Decision Tree (DT), Random Forest (RF), and Artificial Neural Network (ANN) are among the models that are trained after a sequence of procedures to suitably prepare the dataset. In order to wrap up this study, the top-performing model as determined by performance metrics is integrated into a Flask web application. The integration of machine learning models for end-user accessibility is frequently overlooked in existing research, which frequently concentrates only on model performance. This study makes a contribution by outlining the thorough procedures necessary to include the chosen model into an intuitive web application utilizing the Flask framework, in addition to comparing the suggested model's performance with earlier research.

Keywords: Security, Phishing, Conventional based methods, LR, KNN, SVM, DT, RF, ANN, Flask.

| DOI: 10.17148/IARJSET.2021.85125