EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Assessing Digital Government Systems through AI-Powered Techniques

Zaid Ibrahim Rasool

CyberSystem Journal · 24.06.2026 · 3 (1) · 12-20

DOI: 10.57238/csj.2026.1022 ↗

Kurz gefasst

Fachartikel in der Zeitschrift CyberSystem Journal, online erschienen am 24.06.2026. Verfasst von Zaid Ibrahim Rasool. Die Arbeit nennt 25 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Digital government represents the implementation of information and communication technologies (ICT) in provision of public services, governance process optimization and transparent interaction with citizens. However, despite rapid progress in the field, the comprehensive assessment of digital government maturity and quality is rather challenging due to complex nature of related measures. Existing frameworks for evaluation of digital government—from EGDI of UN to DESI of EU—are based on complex indexation across many metrics, thus rendering comparisons difficult and impeding specific recommendations for improvements. This paper proposes a new AI-based approach to digital government assessment and benchmarking based on use of Kohonen Self-Organizing Map (SOM) neural network and focused on the generation of comprehensible cluster maps of digital government performance at national and sub-national level. Data set comprising information over ten-year period was gathered from government repositories and supplemented by public survey with five major dimensions addressed: Data, Technology, Service, People, and Governance. Preprocessing of the data included an elaborate preprocessing pipeline involving the use of imputation, elimination of outliers, data normalization, and principal component analysis for dimensionality reduction before modeling could be undertaken. The results from this process showed four groups of performance, each with built-in outlier detection. Validation of results was achieved using ten-fold cross-validation. The technique proved more effective and accurate than conventional techniques like K-means and DBSCAN and multilayer perceptrons, as indicated by a Davies-Bouldin index value of 0.43 and an accuracy of 93.2%. These results suggest that unsupervised deep learning can be effectively employed as a tool by policymakers and digital government administrators.

Zitieren

APA

Rasool, Z. I. (2026). Assessing Digital Government Systems through AI-Powered Techniques. CyberSystem Journal, 3(1), 12-20. https://doi.org/10.57238/csj.2026.1022

Vancouver

Rasool ZI. Assessing Digital Government Systems through AI-Powered Techniques. CyberSystem Journal. 2026;3(1):12-20. doi:10.57238/csj.2026.1022

BibTeX
@article{Rasool2026,
  author = {Rasool, Zaid Ibrahim},
  title = {Assessing Digital Government Systems through AI-Powered Techniques},
  journal = {CyberSystem Journal},
  year = {2026},
  volume = {3},
  number = {1},
  pages = {12--20},
  doi = {10.57238/csj.2026.1022}
}
Kostenlos starten →