Toward Behavior-Based Adaptive Decision-Making Systems for Cognitive Cities

Smart and cognitive cities rely on extensive human system interaction, making human behavior a critical factor in cybersecurity risk. Existing risk assessment approaches mainly focus on technical vulnerabilities, with limited support for modeling human-centric risks, particularly those influenced by trust and behavior. This paper introduces a behavior-aware architectural mechanism that integrates trustbased human error estimation into runtime system adaptation for cognitive city systems. The proposed framework was developed through a focused literature review of human-factor cybersecurity models, behavioral trust metrics, and human reliability analysis, followed by a conceptual integration of behavior-based trust indicators into the Human Error Assessment and Reduction Technique (HEART). Behavioral trust indicators are derived from observable system interactions rather than self-reported measures, enabling their use as runtime behavioral signals in adaptive cognitive city systems. The paper presents conceptual trust-aware extension of human error estimation that illustrates how behavioral trust can serve as a runtime architectural signal for dynamic system reconfiguration in cognitive cities. By linking trust metrics to adaptive control flows, the framework demonstrates how security controls can self-adjust based on real-time human reliability. The proposed approach provides a conceptual foundation for incorporating human behavioral evidence into system architectures for cognitive cities.