Onto-Security Checks: Semantic Ontology-Based Reasoning for Automated Architectural Security Verification

Abstract

Software architectures are increasingly exposed to security threats that are difficult to detect at the design stage using traditional verification methods. This paper presents Onto-Security Checks, an ontology-based framework for automated security verification at the architectural level. Grounded in OWL (Web Ontology Language), the proposed security ontology models key domain entities users, systems, vulnerabilities,threats, countermeasures,and risks along with their semantic relationships.The theoretical framework relies on knowledge representation and automated reasoning: SWRL rules and SPARQL queries are used to detect unprotected components, infer high-level risk chains from known weaknesses, and recommend appropriate countermeasures. The methodology unfolds across seven phases, from architecture extraction and ontological modeling to automated reasoning and security report generation. The approach was validated through two case studies: the Damn Vulnerable Web Application (DVWA) and a Pharmacy Management System (PMS). Results show a detection rate of 1.00 across all tested scenarios, with reasoning times under 2 seconds for a single system and under 9 seconds for ten replicated subsystems, confirming practical scalability for small to medium-sized architectures.