Northern Kentucky University
A Security-Oriented Awareness and Training Framework using Dynamic and Adaptive Behavior Learning Models
Grade Level at Time of Presentation
Junior
Major
Computer Information Technology, Computer Science
Minor
Computer Science, Math
Institution 24-25
Northern Kentucky University
KY House District #
Campbell County
KY Senate District #
Campbell
Faculty Advisor/ Mentor
Dr. Rasib Khan
Department
Computer Science and Software Engineering
Abstract
The growing cybersecurity challenges across industries often arise from human behavior, such as a lack of cyber awareness or naivety. To address the human factor in security vulnerabilities, we propose a unified model for enhancing security awareness and learning. The model adapts User and Entity Behavior Analytics (UEBA) to collect and analyze behavioral patterns, identifying user interactions with systems and classifying their security behavior. Using a continuous feedback loop, the model updates user behavior by employing diverse methods of deliverance, including gamification, documentation, and training courses. Machine learning and pattern recognition techniques, powered by Large Language Models (LLMs), form the core of the information processing framework. These techniques enhance the system’s ability to dynamically adapt to user needs and foster safer security practices. We posit that this unified approach offers a scalable and effective solution to address and mitigate human vulnerabilities, which are individually unique, and offers a wholistic posture in assuring organizational information security.
A Security-Oriented Awareness and Training Framework using Dynamic and Adaptive Behavior Learning Models
The growing cybersecurity challenges across industries often arise from human behavior, such as a lack of cyber awareness or naivety. To address the human factor in security vulnerabilities, we propose a unified model for enhancing security awareness and learning. The model adapts User and Entity Behavior Analytics (UEBA) to collect and analyze behavioral patterns, identifying user interactions with systems and classifying their security behavior. Using a continuous feedback loop, the model updates user behavior by employing diverse methods of deliverance, including gamification, documentation, and training courses. Machine learning and pattern recognition techniques, powered by Large Language Models (LLMs), form the core of the information processing framework. These techniques enhance the system’s ability to dynamically adapt to user needs and foster safer security practices. We posit that this unified approach offers a scalable and effective solution to address and mitigate human vulnerabilities, which are individually unique, and offers a wholistic posture in assuring organizational information security.