Model-Based Authoring of SOC Simulation Exercises for Cybersecurity Education
Felix Härer
Abstract
Cybersecurity training and classroom exercises with a virtual Security Operations Center (SOC) combine alerts, evidence, timeline entries, learner tasks, and assessment items that must remain consistent across an evolving incident. This paper presents a model-based authoring method in which a technical and educational specification is used to construct an incident exercise model and a large language model (LLM) generates selected scenario text based on that model. The incident exercise model combines an incident model of events, objects, and evidence with an exercise orchestration model of SOC alerts, timeline entries, learner tasks, and assessment items. The method is implemented in a Virtual SOC prototype and evaluated using a phishing-to-ransomware reference exercise. Its incident exercise model contains eight incident events, eight incident objects, six evidence items, seven alerts in five batches, 13 timeline entries including 12 learner tasks, and five representative assessment items. GLM-5.3-generated alert descriptions are inserted and a separate validator checks references, ordering, evidence links, assessment completeness, and selected fixed fields. Validation is demonstrated with reference data, resulting in true-negative outcomes for valid data and true-positive outcomes for deliberately placed inconsistencies. The resulting exercise package is executable within the Virtual SOC learner environment. Overall, the evaluation indicates consistency across the involved model elements and establishes the feasibility of the model-based and generative approach.
Create a lesson
Related papers
SoK: Assessing the Evolution of LLM Capability in Software Vulnerability Research
Amine Lbath, Manan Suri, Aurelien Delaitre et al.
Text-Aware, Vision-Blind: How Multimodal Image Editing Systems Handle Unsafe Inputs
Ziqing Yang, Yixin Wu, Michael Backes et al.
GEOHunter: Mapping a Generative Engine Optimization Network in the Wild
Sean Tang, Peiran Wang, Ying Li et al.
Replication of Agentic Security Events: Audit and Investigation (ROGUE-AI)
Mengxiao Wang, Mahbub Alam, Kalyan Nakka et al.
Backtracking Agent Applications Security Incidents
Yicheng Cai, Taegyu Kim, Peng Liu
Visual Attribution Limits and Graded Provenance Auditing in Commercial Video-Generation APIs
Haocheng Mei, Yuchuan Tian, Yingte Shu et al.