Perancangan Basis Pengetahuan pada Sistem Pakar Identifikasi Jenis Serangan Keamanan Jaringan dengan Metode Rule-Based System

Abstract

Network security threats continue to evolve along with the increasing complexity and scale of network infrastructures. Therefore, a threat analysis model that can represent attack patterns in a structured and systematic manner is required. This study aims to design a knowledge base for network security threat analysis using a rule-based reasoning approach with a forward chaining inference mechanism. Knowledge is represented in the form of IF–THEN rules, which are constructed based on attack characteristics and mapped to the MITRE ATT&CK framework. The developed knowledge base covers ten major types of threats, including attacks on service availability, unauthorized access, malicious code execution, lateral movement, and reconnaissance activities. Knowledge base validation was conducted through theoretical conformity assessment, logical consistency analysis of the rules, and testing using conceptual test scenarios. The results show that all rules produce accurate and consistent inferences according to the given threat conditions, with no conflicts or ambiguities identified. The findings indicate that the rule-based approach with forward chaining is effective for modeling and analyzing network security threats at a conceptual level. The resulting knowledge base model can serve as a foundation for the development of network security threat analysis or detection systems in future research.