Detecting Malicious HTTP Requests Without Log Parser Using RequestBERT-BiLSTM
Web servers provide most internet services, such as information sharing, financial, health, entertainment, and education. In this context, the web has become the principal place for attackers. Unfortunately, most defensive techniques for web servers cannot deal with the complexity and evolution of cyber attacks on HTTP requests. However, machine learning approaches can help detect some attacks. This work presents the RequestBERT-BiLSTM, a new model to detect possible HTTP request attacks without using Log Parser. We evaluated the model on public datasets such as CSIC 2010, ECML/PKDD 2007, and BGL. We also developed a new dataset from a real environment to evaluate the method. In addition, we illustrate that the traditional log analysis step can degrade the model’s performance due to parser errors. Furthermore, we compared the performance of the proposed approach with literature models, and we obtained a detection rate above 95%.
- Identification of Microorganism Colony Odor Signature using InceptionTime
- Intrusion Detection for Cyber–Physical Systems Using Generative Adversarial Networks in Fog Environment
- LogBERT-BiLSTM: Detecting Malicious Web Requests
- Multi-Class Mobile Money Service Financial Fraud Detection by Integrating Supervised Learning with Adversarial Autoencoders
- PictoBERT: Transformers for Next Pictogram Prediction