Abstract
The construction of an for with dynamic, flexible, and intelligent features is a new idea. To solve the problem whereby traditional static protection methods cannot respond to various network attacks or security demands in an adversarial network environment in time, and to form a complete from “threat discovery” to “decision-making generation,” we propose an -based security model, OntoCSD, for an of that uses Web language (OWL) to represent the classes and relationships of threat monitoring, decision-making, response, and defense in cyberspace, and uses semantic Web rule language (SWRL) to design the defensive reasoning rules. OntoCSD can discover potential relationships among network attacks, vulnerabilities, the security state, and defense strategies. Further, an artificial intelligence (AI) expert system based on is used to quickly generate a detailed and comprehensive decision-making scheme. Finally, through Kendall’s coefficient of concordance () and four experimental cases in a typical system, which reasons on represented facts and the , OntoCSD’s consistency and its feasibility to solve the issues in the field of are validated. OntoCSD supports automatic association and reasoning, and provides an framework of .