Cybercrime Victim Profiling in Nigeria Using Machine Learning and Psychological Traits

Abstract

Cybercrime victimization is on the rise, yet most existing studies focus on attackers rather than victims. This research examines the role of psychological traits in predicting cybercrime victimization in Nigeria using machine learning techniques. The research is motivated by the need to integrate human behavioral factors into cybersecurity, the study employs Random Forest, Decision Tree, Naïve Bayes, and Logistic Regression models to analyze thelinks between the Big Five personality traits and victim susceptibility. Data was collected through a SurveyMonkey questionnaire administered to residents of Abuja Municipal Area Council (AMAC) and a secondary dataset from an open-access Big Five personality repository. The models were trained and evaluated using accuracy, precision, recall, and F1 score metrics after data preprocessing. Random Forest achieved the highest accuracy at 97.2%. From our findings, individuals with high extraversion and low agreeableness, conscientiousness, emotional stability, and openness are more vulnerable to cybercrime. These insights support the development of personality-informed cybersecurity awareness and prevention strategies.