Machine Learning-based web security intrusion detection system
Chen Chuanlin, Jing Zhong, Wenjie Chen · 2021 3rd International Academic Exchange Conference on Science and Technology Innovation (IAECST) · 2021
In recent years, the rise of technologies such as dynamic web pages have led to the emergence of many web applications. In addition, the complexity and variety of attacks against web applications make it difficult for intrusion detection systems based on misuse detection to use constructed attack signatures for detection effectively. This paper first presents two common web attacks, SQL injection and XSS cross-site scripting attacks, their specific classification, followed by a simple optimization of the collected SQL and XSS datasets, the optimization of the model parameters using machine learning algorithms, and a comparison of the performance of the different algorithms on GPU/CPU. In addition, they were deployed in a production environment for practical testing. After experimental results and error analysis, the best algorithmic model was determined. Based on these experiments, a machine learning-based network intrusion detection system was designed without compromising the quality of the webserver.