AN EXPLAINABLE MACHINE LEARNING FRAMEWORK FOR CYBERSECURITY AND DIGITAL SELF GOVERNANCE IN LOCAL GOVERNMENT NETWORKS

Avtorji

  • R Kalaivani, Dr. V.Vennila

DOI:

https://doi.org/10.52152/zsc9dh53

Ključne besede:

Digital Self-Governance; Intrusion Detection; Explainable Machine Learning; Federated Learning; Local Government Cybersecurity.

Povzetek

Local governments increasingly depend on connected service platforms, networked sensors, digital records, and inter-organizational data exchange. These systems create an operational need for intrusion detection that is accurate, explainable, privacy-conscious, and compatible with the accountability obligations of local self-government. This article proposes a governance-aware machine-learning framework that combines local network monitoring, privacy-aware preprocessing, lightweight supervised detection, explainable alerts, and federated model improvement. The technical core specifies data preparation, class-imbalance handling, model training, risk scoring, and evaluation through six formal expressions. A governance control plane links model outputs to human verification, role-based access, retention rules, audit logs, and incident-response responsibilities. The design is intended for municipal networks with heterogeneous technical capacity and limited ability to centralise sensitive traffic data. A reproducible evaluation protocol is defined around public intrusion-detection benchmarks, time-aware train-validation-test separation, and accuracy, precision, recall, F1-score, false-positive rate, latency, and communication cost. To demonstrate how the proposed evaluation should be reported, the manuscript includes an explicitly illustrative numerical scenario; these values are not empirical findings and must be replaced by results from executed experiments before submission. The contribution is therefore a testable engineering architecture coupled with a local-governance implementation model, rather than a claim that a deployed municipal system has already been validated.

Objavljeno

2026-09-01

Številka

Rubrika

Article

Kako citirati

AN EXPLAINABLE MACHINE LEARNING FRAMEWORK FOR CYBERSECURITY AND DIGITAL SELF GOVERNANCE IN LOCAL GOVERNMENT NETWORKS. (2026). Lex Localis - Journal of Local Self-Government, 313-326. https://doi.org/10.52152/zsc9dh53