Securing Web Applications Against Payload Attacks Using Deep Learning-Based Model

Source abstract

Web applications are critical to modern systems but are increasingly targeted by payload attacks, where malicious data is injected to exploit vulnerabilities, steal information, or disrupt operations. SQL injection and cross-site scripting (XXS) attacks are two of the most common types of attacks that can compromise web applications. This paper proposes a model for the detection of SQL and XXS attacks, which aims to enhance the security of web applications and prevent potential damages. The methodology involves three main stages: data preprocessing, feature extraction, and model training and classification. First, a dataset consisting of structured queries, which comprise of both safe and unsafe (XSS and SQL injection attack) queries, is collected and preprocessed. Then, features are extracted from the dataset using statistical and machine learning techniques, which are used to train and test LSTM-based classification model. Finally, the model is evaluated using different performance metrics, including accuracy, precision, recall, and F1-score. The results showed that the proposed model can effectively detect SQL and XXS attacks with a high accuracy rate of 98%, and can also handle different attack scenarios with low false positive rates.