Skip to Main Content

Modelling and Data Analysis
2026. Vol. 16, no. 1, 50–60
https://doi.org/10.17759/mda.2026160103
ISSN: 2219-3758 / 2311-9454 (online)

Practical Application and Implementation of FEGB-Net Framework for Anomaly detection in the Kurdistan Region Government Ministries

 
Audio is AI-generated
46

PDF (RUS)
Online translation in english
GS
Information
in Google Scholar
Metrics

Abstract

This paper presents the deployment and application of the Federated Ensemble Graph-Based Network (FEGB-Net) framework within the Kurdistan Region Government (KRG) ministries. The system integrates Federated Learning (FL), Graph Neural Networks (GNNs), and ensemble machine learning to provide privacy-preserving and collaborative anomaly detection in distributed government networks. Real-world deployment across key ministries demonstrated improved detection accuracy (97.6 %), low false-positive rates (3.2 %), and enhanced resilience against adversarial and stealthy attacks, while maintaining full compliance with governmental data-sovereignty requirements.

General Information

Keywords: machine learning, intrusion detection system (IDS), Federated Learning (FL), graph neural networks (GNN), ensemble learning, privacy, confidentiality, security, digital government

Journal rubric: Data Analysis

OpenAlex citations: 0

OpenAlex trends: Advanced Graph Neural Networks, Privacy-Preserving Technologies in Data, Big Data and Digital Economy

Information about the work in OpenAlex

Number of citations: 0

Topics

Advanced Graph Neural Networks

This cluster of papers focuses on the development, applications, and techniques related to Graph Neural Networks (GNNs) and their variants. It covers topics such as knowledge graph embedding, representation learning, network embedding, deep learning, graph convolutional networks, heterogeneous networks, relational data modeling, signal processing on graphs, and semi-supervised learning.

Number of works: 59136  |  Total number of citations: 573016

Topic detailsв OpenAlex

Privacy-Preserving Technologies in Data

This cluster of papers focuses on privacy-preserving techniques for data analysis and machine learning, including topics such as differential privacy, federated learning, k-anonymity, secure computation, and location privacy. The papers explore methods to protect sensitive information while performing data mining, machine learning, and statistical analysis.

Number of works: 84970  |  Total number of citations: 1083596

Topic detailsв OpenAlex

Big Data and Digital Economy

This cluster of papers focuses on the intersection of big data and cloud computing technologies, with an emphasis on security, energy efficiency, machine learning, IoT, privacy protection, resource allocation, and cybersecurity in the context of mobile sensing. The papers cover a wide range of topics including data optimization, secure communication, efficient resource allocation, and privacy-preserving strategies.

Number of works: 48434  |  Total number of citations: 44406

Topic detailsв OpenAlex

Work details in OpenAlex

Article type: scientific article

DOI: https://doi.org/10.17759/mda.2026160103

Acknowledgements. The author would like to thank E.V. Lyapuntsova, Doctor of Technical Sciences, Professor, for valuable advice in planning the study and discussing the results.

Received 17.12.2025

Revised 19.01.2026

Accepted

Published

For citation: Arm, A.A.S. (2026). Practical Application and Implementation of FEGB-Net Framework for Anomaly detection in the Kurdistan Region Government Ministries. Modelling and Data Analysis, 16(1), 50–60. (In Russ.). https://doi.org/10.17759/mda.2026160103

© Arm A.A.S., 2026

License: CC BY-NC 4.0

References

  1. Арм А.А.С., Ляпунцова Е.В. Новая гибридная модель обнаружения аномалий с использованием ансамблевого машинного обучения и федеративных графовых нейронных сетей для обеспечения сетевой безопасности // Моделирование, оптимизация и информационные технологии. Т. 13, № 2. DOI: 10.26102/2310-6018/2025.49.2.044.
    Arm A.A.S., Lyapuntsova E.V. A novel hybrid anomaly detection model using federated graph neural networks and ensemble machine learning for network security. Modeling, Optimization and Information Technology. 2025;13(2). (In Russ.). DOI: 10.26102/2310-6018/2025.49.2.044
  2. Ahmad R., et al. Hybrid CNN-LSTM intrusion detection // Applied Intelligence. 2022. Vol. 52. P. 10013–10027. DOI: 10.1007/s10489-021-02866-z.
  3. Ahmad R., Shamsuddin K. A systematic literature review of intrusion detection systems for IoT networks // IEEE Access. 2021. Vol. 9. P. 5784–5810. DOI: 10.1109/ACCESS.2021.3050346.
  4. Chaabane A., et al. Cyberattack categorization and defense mechanisms in government digital services // Government Information Quarterly. 2022. Vol. 39, No. 3. DOI: 10.1016/j.giq.2022.101696.
  5. Chen Y., et al. Hybrid deep-learning architectures for intrusion detection // Computers & Security. 2023. Vol. 126. Article 103046. DOI: 10.1016/j.cose.2023.103046.
  6. Coull S. E., Teng T. H. Detecting insider threats using user-activity graph modelling // IEEE Access. 2020. Vol. 8. P. 185351–185365. DOI: 10.1109/ACCESS.2020.3029429.
  7. Dwork C., et al. Differential privacy: A survey of results // ACM Computing Surveys. 2022. Vol. 54, No. 2. P. 1–38. DOI: 10.1145/3317432.
  8. Javaid A., et al. Comprehensive evaluation of ML-based intrusion detection // IEEE Access. 2021. Vol. 9. P. 102721–102736. DOI: 10.1109/ACCESS.2021.3098461.
  9. Kaspersky N., et al. Early-stage ransomware detection using behavior graphs // Computers & Security. 2022. Vol. 123. Article 102930. DOI: 10.1016/j.cose.2022.102930.
  10. Kim J., et al. Benchmarking deep learning models for intrusion detection // IEEE Access. 2023. Vol. 11. P. 8280–8292. DOI: 10.1109/ACCESS.2023.3240121.
  11. Kumar M., Yadav P. Phishing detection using email graph embeddings // Expert Systems with Applications. 2023. Vol. 224. Article 119902. DOI: 10.1016/j.eswa.2023.119902.
  12. Li M., Huang T., Chen Y. Federated learning for network intrusion detection in IIoT: A comprehensive study // IEEE Internet of Things Journal. 2022. Vol. 9, No. 10. P. 7413–7427. DOI: 10.1109/JIOT.2021.3136928.
  13. Li Y., et al. Temporal graph learning for APT detection // IEEE Transactions on Information Forensics and Security. 2023. Vol. 18. P. 1098–1112. DOI: 10.1109/TIFS.2023.3236209.
  14. McMahan B., et al. Communication-efficient learning of deep networks from decentralized data // Proc. AISTATS. 2017.
  15. Nguyen T., et al. Autoencoder-based intrusion detection // IEEE Access. 2021. Vol. 9. P. 17710–17725. DOI: 10.1109/ACCESS.2021.3053265.
  16. Nguyen T., Le M. Graph-based correlation for cyber flow analysis // Journal of Cybersecurity Engineering. 2021. Vol. 8, No. 2. P. 87–98. DOI: 10.1061/JCEITR.0000459.
  17. Zero Trust Architecture for Government Networks. NIST SP 800-207. 2020. DOI: 10.6028/NIST.SP.800-207.
  18. Santos A., et al. Limitations of signature-based IDS and benefits of behavior-based detection // Computers & Security. 2022. Vol. 120. Article 102770. DOI: 10.1016/j.cose.2022.102770.
  19. Wang P., et al. Transformer-IDS: Attention-based intrusion detection // IEEE Transactions on Neural Networks and Learning Systems. 2024. Vol. 35, No. 5. P. 5784–5797. DOI: 10.1109/TNNLS.2023.3254776.
  20. Xia F., et al. Robust federated ensemble learning against model poisoning // IEEE Access. 2023. Vol. 11. P. 45567–45579. DOI: 10.1109/ACCESS.2023.3265831.
  21. Yang Q., et al. Scaling federated learning for intrusion detection under non-IID conditions // IEEE Internet of Things Journal. 2023. Vol. 10, No. 5. P. 4330–4342. DOI: 10.1109/JIOT.2022.3204463.
  22. Zhou S., et al. Graph neural networks for large-scale intrusion detection // IEEE Transactions on Neural Networks and Learning Systems. 2023. Vol. 34, No. 2. P. 912–924. DOI: 10.1109/TNNLS.2021.3139072.

Information About the Authors

Azhi A. Arm, Graduate Student, the Department of Computer-Aided Engineering and Design, National University of Science and Technology "MISIS", Moscow, Russian Federation, e-mail: arm.azhi@yandex.com

Contribution of the authors

Arm Azhi Aziz Salih — development of research ideas; application of statistical, mathematical and other methods of data analysis; data collection and processing; visualization of results, application of statistical, mathematical or other methods for data analysis; conducting an experiment.

Conflict of interest

The authors declare no conflict of interest.

Metrics

 Web Views

Whole time: 182
Previous month: 35
Current month: 13

 PDF Downloads

Whole time: 46
Previous month: 5
Current month: 3

 Total

Whole time: 228
Previous month: 40
Current month: 16