Web application vulnerability testing framework

Abstract

The increasing prevalence of cyberattacks targeting web applications necessitates advanced vulnerability detection techniques. Traditional methods such as Static Application Security Testing (SAST) and Dynamic Application Security Testing (DAST) face challenges including high false positives, limited coverage of modem architectures (e.g., serverless, microservices), and inefficiency in identifying zero-day vulnerabilities. This paper proposes a hybrid vulnerability testing framework that combines SAST, DAST, and Machine Learning (ML) to enhance detection accuracy and adaptability. The technique integrates static code analysis for identifying insecure coding patterns, dynamic runtime monitoring to detect exploitation attempts, and an ML classifier trained on anomaly datasets to reduce false alarms.