DOCTORAL CANDIDATE
Sara Pelinku

SUPERVISORS
Kavé Salamatian, University of Savoie Mont Blanc (USMB)
Olaf Maennel, The University of Adelaide (UoA)
Kaie Maennel, The University of Adelaide (UoA)

Computer science, cyber security, heterogenous data, data visualisation, artificial intelligence, machine learning

The research addresses the critical challenge of integrating heterogeneous cybersecurity data sources into coherent and actionable systems. Information from diverse formats and origins, such as Intrusion Detection Systems (IDS), threat intelligence reports, logs, and SIEMs, will be represented using graph-based methodologies. The proposed system uses multi-graph representations to enable the application of graph mining techniques and multi-criteria optimization, facilitating the fusion of relevant data into a comprehensive decision framework. This approach enhances visualization capabilities, potentially incorporating advanced tools like 3D visualizations for rapid threat analysis and mitigation.

A key use case focuses on highly automated maritime vessels with minimal crew, where technical systems detect navigation or operational anomalies caused by cyberattacks. The graph-based system will transform raw IDS data into intuitive visual formats to ensure seafarers, even without technical expertise, can quickly understand and respond to threats, especially in adversarial conditions that restrict communication. The project aims to develop a system for timely alerts, improve situational awareness through intuitive visualizations, and evaluate threat communication methods for non-technical users.