Nuclear Facility Access Control Based on Bio-signals to Prevent a Potential Insider Threat

(A) Through the literature review, this study will review the state of the art insider threat scenarios and select scenarios that are more likely to occur in Republic of Korea. Based on the suggested scenarios, this study will propose a feasible experimental design. More realistic scenarios will be developed to represent insider threats and corresponding simulation exercises. A criminal psychology expert advisory meeting will be held to understand the concept related to the insider threat. The proposed insider threat scenarios and experimental designs will be modified through consulting with expert advisory.(B) Subjects will be recruited for the insider threat experiment. The subjects will be students whose major are nuclear engineering and nuclear power plant workers. In order to determine the number of subjects, this study will used statistical method. During the experiment, we will record the EEG of the subjects. Groups will be statistically categorized that are more likely to be insiders and less likely to be insiders. Through analyze the EEG of both groups, this study will obtain indicators to find group that will be more likely to be insiders. A brain science expert advisory meeting will be held to understand how to analyze and interpret the results. Finally, the experiment results will be modified through consulting with brain science expert advisory.(C) A machine learning expert advisory meeting will be held to understand the basic concepts of machine learning and select a tool appropriate for insider threat identification. Through the experiment and analysis, this study will propose a model that predicts the group that will be more likely to be insiders using the proposed indicators above.