Assessing cybersecurity awareness and behavioural risk profiles at an aviation-focused University in Abuja, Nigeria, using machine learning segmentation

Exposure to cybersecurity threats and human-related vulnerabilities have been brought about by the digitalization of aviation-relatedinstitutions of learning. Even though these institutions have cybersecurity awareness initiatives, unsafe cyber practices still remain amajor contributor toward institutional cyber-risks. This study assessed cybersecurity awareness and behavioural risk profiles at AfricanAviation and Aerospace University (AAAU), Abuja, using behavioural cybersecurity modelling and machine-learning-basedsegmentation techniques. The survey was conducted electronically using Google forms on respondents across different sections of theUniversity. 151 valid responses were obtained from academic and non-academic staff, undergraduate and postgraduate students andinterns. Cybersecurity Risk Index (CRI) modelling, awareness-behaviour gap analysis, comparative institutional analysis, descriptivestatistics, reliability analysis and K-means clustering were employed. Findings revealed that respondents demonstrated moderate-tohigh cybersecurity awareness levels even though several respondents engaged in risky cybersecurity activities like password reuse, unsafe Wi-Fi usage, and delayed software update. 50.99% of respondents fell within the moderate-risk category on the CRI analysis while Kmeans clustering identified 3 profiles: Aware-but-Vulnerable Users (23.84%), High-Risk/Low-Awareness Users (33/11%) and SecurityConscious Users (43.05%). Results show that cybersecurity awareness alone is not sufficient to predict a secure cybersecurity behaviour. A Context-Aware Cybersecurity Framework for Aviation Education (CACF-AEI) was proposed based on the result and it integrated behavioural cybersecurity profiling, contextual cybersecurity indicators, adaptive cybersecurity interventions and machine-learning segmentation. The study contributes to behavioural cybersecurity analytics through the integration of risk modelling and unsupervised machine learning within an aviation-focused institutional environment.