CyberGuard: A Deployable Deep Learning Model for Cyber Attacks Threat Detection, and Prevention
Document Type
Research Project
Publication Date
Spring 3-17-2026
Abstract
Cyber threats have been exponentially increased by recent advancements in technology, and the vulnerabilities are being exploited by malicious actors to manipulate, steal, destroy or disclose private information. Web applications, especially the ones with sensitive data, are the main targets, and they are continuously attacked through SQL Injection (SQLi), Cross-Site Scripting (XSS), and Phishing. It is not easy to detect these attacks owing to their constantly changing and unclear nature. The present study makes use of deep learning with Long Short-Term Memory (LSTM) and Bidirectional LSTM (BiLSTM) networks, which are trained on three labeled datasets for SQLi, XSS, and phishing attacks, helping the model to grasp complex temporal dependencies and contextual cues. The end objective is to turn this model into a complete security framework that can protect electronic systems holding confidential information.
Recommended Citation
Fahmy, Ahmed M. and Wafiq, Mohamed M., "CyberGuard: A Deployable Deep Learning Model for Cyber Attacks Threat Detection, and Prevention" (2026). Artificial Intelligence. 27.
https://buescholar.bue.edu.eg/artificial_intelligence/27