Performance Evaluation of Machine Learning Algorithms for Website Defacement Attack Detection

In recent years, Website defacement is a common attack on web servers, when a hacker changes the content of a website to serve their own purposes. Detecting such defacements is crucial for maintaining the integrity and availability of web content. In this study, we explore the effectiveness of four machine learning algorithms, Decision Tree, Random Forest, Stochastic Gradient Descent, and Extra Trees, for detecting website defacement attacks. We evaluate the performance of these algorithms on a publicly available dataset of defaced and non-defaced websites. Our results show that Extra Trees outperforms the other three algorithms in terms of accuracy, precision, recall, and F1 score. Extra Trees is able to achieve an accuracy of 89%, a precision of 90%, a recall of 95%, and an F1 score of 94%. We attribute the superior performance of Extra Trees to its ability to create multiple decision trees and randomize the selection of features. This enables it to capture more complex relationships between the features and the target variable, leading to better generalization and lower overfitting. Overall, our findings suggest that Extra Trees is a highly effective algorithm for website defacement detection and can be a valuable tool for enhancing the security of web servers.