SCVI: Bridging Social and Cyber Dimensions for Comprehensive Vulnerability Assessment

Abstract

The increasing prevalence of cyber threats on social media demands advanced metrics for assessing and mitigating social cyber vulnerabilities. This thesis introduces the Social Cyber Vulnerability Index (SCVI), a novel framework that integrates individual-level factors (e.g., awareness, behavioral traits, psychological attributes) with attack-level characteristics (e.g., frequency, impact, sophistication) for a comprehensive socio-cyber vulnerability assessment. Validated using survey data (iPoll) and textual data (Reddit scam reports), SCVI demonstrates adaptability across modalities while identifying demographic and regional disparities. Comparative analyses with the Common Vulnerability Scoring System (CVSS) and the Social Vulnerability Index (SVI) highlight SCVI's superior sensitivity to socio-technical risks. Monte Carlo-based weight variability analysis confirms its robustness and effectiveness in identifying high-risk groups. By addressing limitations in existing metrics, SCVI provides actionable insights for policymakers and practitioners, enhancing strategies to counter emerging threats like AI-driven phishing and deepfake scams.