A deep learning-based adaptive cyber disaster management framework

A deep learning-based adaptive cyber disaster management framework

연구 분야: Safety

학회: Journal of Big Data

The prevalence of cybersecurity incidents targeting Industrial Control Systems (ICS) in critical national infrastructure sectors has alarmingly risen. Given ICS’s cyber-physical nature, cyber incidents, directly and indirectly, affect public health, social stability, environmental quality, and national security, introducing a new type of humanitarian threat: cyber disaster. It is generally accepted that novel data-driven methodologies must address diverse physical and societal challenges from man-made hazards to protect human well-being and social welfare. We argue that evolving cyber threats in the context of critical infrastructure require a novel approach that treats cyber incidents as disasters and leverages advances in deep learning techniques, fully integrating them into the disaster management cycle. Our argument adds to the ongoing conversation about improving disaster management practices to address contemporary challenges. In this study, we offer a deep learning-based approach to cyber disaster management. We draw upon disaster management cycles, deep learning, and new cyber threat detection instantiation. Our study offers both practical and methodological contributions to the knowledge base for scholars seeking to leverage deep learning to design deep learning-based approaches needed to thwart increasingly sophisticated cyber incidents.