Deep Learning-Based Phishing URL Detection Using Deep Neural Network and Convolutional Neural Networks

Phishing attacks remain one of the most significant cybersecurity threats, where malicious websites attempt to deceive users into revealing sensitive information such as login credentials and financial data. Traditional phishing detection techniques often struggle to identify newly generated phishing websites due to the rapidly evolving tactics used by attackers. This study proposes a deep learning-based approach for phishing URL detection by comparing the performance of two neural network architectures, namely a Deep Neural Network (DNN) and a one-dimensional Convolutional Neural Network (CNN). The experiments were conducted using a phishing URL dataset containing structural and behavioral features extracted from website and URL characteristics. Data preprocessing was performed by removing irrelevant identifiers, eliminating duplicate samples, and applying feature normalization to improve model training stability. The dataset was divided into training, validation, and testing subsets using stratified sampling to preserve the class distribution. Both models were evaluated using multiple performance metrics including accuracy, precision, recall, F1 score, and the area under the receiver operating characteristic curve. The experimental results show that the CNN model achieved superior performance with an accuracy of 93.2 percent and an AUC value of 0.988, outperforming the DNN model across all evaluation metrics. These findings demonstrate that convolutional neural networks are more effective in capturing complex feature interactions within phishing URL datasets and can provide a reliable approach for automated phishing detection systems in cybersecurity applications.