XAI-driven Data Mining for Self-defending IoT Systems: Enhancing Cybersecurity Transparency in the Age of Smart Cities

The rapid expansion of Internet of Things (IoT) technologies in smart cities, healthcare, and industrial automation has intensified the need for cybersecurity frameworks capable of operating at scale and in real time under increasingly sophisticated threat conditions. Traditional security mechanisms and opaque AI-based models are no longer adequate for protecting interconnected urban infrastructures, especially as regulatory and societal expectations move toward transparency and accountability. Although prior surveys have examined IoT security and general AI techniques, they rarely address the emerging role of Explainable Artificial Intelligence (XAI) in data mining for IoT cybersecurity or integrate recent advances in cognitively inspired and human-aligned explainability methods. This survey provides an up-to-date review of XAI-driven data mining approaches applied to IoT ecosystems, highlighting their ability to detect anomalies, interpret complex sensor-driven behaviours, and support automated security decisions through transparent and interpretable reasoning. The review identifies critical challenges, including data privacy, scalability, computational constraints, and the interpretability limitations of modern AI models. It examines how biologically inspired learning paradigms and cognitively grounded explanation techniques can enhance trust and situational awareness in IoT environments. Emerging technologies such as edge intelligence, federated learning (FL), blockchain integration, and quantum-assisted analytics are discussed as promising enablers of scalable and transparent IoT security. The survey underscores the importance of trustworthy, ethically aligned AI, advocating for XAI frameworks that enable fair, auditable, and reliable decision-making in safety-critical infrastructure. By addressing gaps in the literature and synthesizing recent developments, this study presents a timely perspective on XAI, data mining, and IoT cybersecurity, outlining future directions for building resilient, interpretable, and human-centric smart city systems.