XSS attack detection based on machine learning

• Autor(a) principal: Shan, Waner
• Data de Publicação: 2023
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10362/153183
• topic: Cross-site scripting; supervised learning algorithms; classifiers; javascript; DOM; HTTP; Domínio/Área Científica::Engenharia e Tecnologia::Engenharia Eletrotécnica, Eletrónica e Informática

As the popularity of web-based applications grows, so does the number of individuals who use them. The vulnerabilities of those programs, however, remain a concern. Cross-site scripting is a very prevalent assault that is simple to launch but difficult to defend against. That is why it is being studied. The current study focuses on artificial systems, such as machine learning, which can function without human interaction. As technology advances, the need for maintenance is increasing. Those maintenance systems, on the other hand, are becoming more complex. This is why machine learning technologies are becoming increasingly important in our daily lives. This study use supervised machine learning to protect against cross-site scripting, which allows the computer to find an algorithm that can identify vulnerabilities. A large collection of datasets serves as the foundation for this technique. The model will be equipped with functions extracted from datasets that will allow it to learn the model of such an attack by filtering it using common Javascript symbols or possible Document Object Model (DOM) syntax. As long as the research continues, the best conjugate algorithms will be discovered that can successfully fight against cross-site scripting. It will do multiple comparisons between different classification methods on their own or in combination to determine which one performs the best.