Securing web applications against XSS and SQLi attacks using a novel deep learning approach.

Area of Science:

• Cybersecurity
• Machine Learning
• Web Application Security

Background:

• Web applications handle sensitive data, making security paramount.
• SQL injection (SQLi) and cross-site scripting (XSS) are common, damaging attacks.
• Existing detection methods often have high false positive rates and limited accuracy.

Purpose of the Study:

• To propose a novel approach for securing web applications against SQLi and XSS attacks.
• To enhance the accuracy and reduce false positives in attack detection.
• To leverage hybrid deep learning for improved pattern recognition in security threats.

Main Methods:

• Developed a hybrid deep learning (DL) model combining Convolutional Neural Networks (CNNs) and Long Short-Term Memory (LSTM) networks.
• Employed decoding and standardization of SQL/XSS payloads and HTTP requests.
• Trained and evaluated the model on custom testbed, SQL/XSS payload, and HTTP CSIC 2010 benchmark datasets.

Main Results:

• Achieved high accuracy rates: 99.84% (SQL-XSS Payload), 99.23% (Testbed), and 99.77% (HTTP CSIC 2010).
• Demonstrated superior performance compared to traditional machine learning methods.
• Effectively identified SQLi and XSS attacks with a low false positive rate.

Conclusions:

• The proposed hybrid DL model offers an effective solution for detecting SQLi and XSS attacks.
• The approach shows promise for broader network security applications like intrusion detection systems and web application firewalls.
• Deep learning techniques enhance the ability to detect subtle attack patterns missed by conventional methods.