Cyber-Attack Detection of Malicious URLs Using Deep-Learning Techniques

The rapid growth of internet usage in daily life has led to a significant increase in cyber threats, with malicious URLs being a common vector for cybercrime. Traditional detection methods often exhibit high false-alarm rates and struggle to keep pace with evolving threats, owing to outdated feature-extraction techniques and datasets. To address these limitations, a deep learning-based approach has been proposed aimed at developing an effective model for detecting malicious URLs. Our proposed method focused on the char2B model, which leverages a fusion of BERT and CharBiGRU embedding, further enhanced by a Conv1D layer with a kernel size of 3 and unit-sized stride and padding. After combining the embeddings, the BERT model served as the baseline for comparison. The study involved collecting a dataset of 87,216 URLs, comprising both benign and malicious samples, sourced from the Open Directory Project (DMOZ), PhishTank, and Any.Run. Models were trained using the training set and evaluated on the test set using standard metrics, including accuracy, precision, recall, and F1 score. Through iterative refinement, we optimized the model’s performance to maximize its effectiveness. As a result, our proposed model achieved 98.50% accuracy, 98.27% precision, 98.69% recall, and 98.48% F1 score, outperforming the baseline BERT model. Additionally, the false positive rate of our model was 0.017 better than the baseline model’s 0.018. By effectively extracting and using informative features, the model accurately classified URLs as benign or malicious, thereby improving detection performance. This study highlights the significance of our deep learning approach in strengthening cybersecurity by integrating advanced algorithms that enhance detection accuracy, strengthen defenses, and contribute to a safer digital environment.