Smart Cities are more digitalized and dependent on interconnecting digital infrastructures and e-Governance platforms that provide efficient public services hence becoming very susceptible to advanced cyber threats. The need to have strong cybersecurity and at the same time have uninterrupted governance operations have made this a critical issue. This article suggests an AI-based cybersecurity and e-Governance system, which combines a machine learning and ensemble learning model to improve threat detection, risk evaluation, and governance decision support in the Smart City setting. The proposed methodology integrates heterogeneous traffic data of Smart City network with records, which are related to governance and then preprocessed, feature engineered, and anonymized by preserving privacy. There are several learning models which are used to classify cyber threats and anomalies in governance such as the Artificial Neural Networks, Support Vector Machines, Logistic Regression, Decision trees and the Gradient Boosting. The validation of the experiment is performed based on the hybrid data set consisting of simulated e-Governance data and example cybersecurity traffic rates. Accuracy, precision, recall, and false positive rate are used to determine performance. The findings indicate that Gradient Boosting and ANN are superior to the conventional models as they are more accurate in detection and lower false alarm, making them reliable and practicable in use. The results prove that the developed AI-based framework contributes to the cybersecurity resiliency to a large extent and promotes e-Governance in Smart City ecosystems based on transparent and data-driven results.
Vidyasagar et al. (Thu,) studied this question.