1. MMC
  2. All Volumes and Issues
  3. Volume 12, Number 3, 2025
  4. Comparative analysis of networks' centrality measures with ANOVA

Comparative analysis of networks' centrality measures with ANOVA

MMC.
2025;
Volume 12, Number 3
: pp. 809–818
https://doi.org/10.23939/mmc2025.03.809
Received: February 09, 2025
Revised: August 29, 2025
Accepted: August 30, 2025

Mukhtar M. F., Khashi'ie N. S., Nordin S. K. S., Zainal N. A., Abas Z. A.  Comparative analysis of networks' centrality measures with ANOVA.  Mathematical Modeling and Computing. Vol. 12, No. 3, pp. 809–818 (2025)

Authors:
  1. M. F. Mukhtar
  2. N. S. Khashi'ie
  3. S. K. S. Nordin
  4. N. A. Zainal
  5. Z. A. Abas
1
Faculty of Mechanical and Manufacturing Engineering Technology, Universiti Teknikal Malaysia Melaka
2
Faculty of Mechanical and Manufacturing Engineering Technology, Universiti Teknikal Malaysia Melaka
3
Faculty of Mechanical and Manufacturing Engineering Technology, Universiti Teknikal Malaysia Melaka
4
Faculty of Mechanical and Manufacturing Engineering Technology, Universiti Teknikal Malaysia Melaka
5
Faculty of Information and Communication Technology, Universiti Teknikal Malaysia Melaka

This study introduces the GDK method, combining Global Structure Model (GSM), Degree Centrality (DC), and K-shell decomposition (Ks), to assess node significance in networks.  In comparison to traditional metrics (Degree Centrality, Betweenness Centrality, and Closeness Centrality), GDK is evaluated across three network types: social (Email), scientific (Netscience), and technological (Router).  Analysis of Variance (ANOVA) and Kendall's correlation show that GDK consistently achieves higher correlation in ranking nodes, making it a more reliable tool.  By integrating local and global centrality features, GDK identifies key nodes with both direct and structural importance, outperforming single-dimension measures.  For example, in the Email network, GDK highlights both direct and bridging nodes, while in Netscience, it combines local and structural criteria to find influential nodes.  The results suggest that GDK offers a more nuanced evaluation of node importance, addressing the limitations of traditional methods.  Further research should explore its application to larger and more diverse networks.

network
centrality
ANOVA
combinations
  1. Saqr M., López-Pernas S.  The Curious Case of Centrality Measures: A Large-Scale Empirical Investigation.  Networks in Learning Analytics.  9 (1), 13–31 (2022).
  2. D'Avolio E., Bandinelli R., Rinaldi R.  Product development and PLM performance measures: A multiple-case study in the fashion industry.  IFIP Advances in Information and Communication Technology.  492, 399–410 (2016).
  3. Wang B., Zhang J., Dai J., Sheng J.  Influential nodes identification using network local structural properties.  Scientific Reports.  12 (1), 1833 (2022).
  4. Namtirtha A., Dutta A., Dutta B., Sundararajan A., Simmhan Y.  Best influential spreaders identification using network global structural properties.  Scientific Reports.  11 (1), 2254 (2021).
  5. Hallquist M. N., Wright A. G. C., Molenaar P. C. M.  Problems with Centrality Measures in Psychopathology Symptom Networks: Why Network Psychometrics Cannot Escape Psychometric Theory.  Multivariate Behavioral Research.  56 (2), 199–223 (2021).
  6. Sadek A. H., Shepley M. M. C.  Space Syntax Analysis: Tools for Augmenting the Precision of Healthcare Facility Spatial Analysis.  Health Environments Research & Design Journal.  10 (1), 114–129 (2016).
  7. Saxena A., Malik V., Iyengar S. R. S.  Estimating the Degree Centrality Ranking of a Node.  Preorint arXiv:1511.05732v2 (2015).
  8. Bader D. A., Kintali S., Madduri K., Mihail M.  Approximating betweenness centrality.  Algorithms and Models for the Web-Graph.  124–137  (2007).
  9. Yang R., Zhuhadar L.  Extensions of closeness centrality?  ACMSE '11: Proceedings of the 49th annual ACM Southeast Conference. 304–305 (2011).
  10. Wang J., Rong L., Guo T.  A new measure of node importance in complex networks with tunable parameters.  2008 4th International Conference on Wireless Communications, Networking and Mobile Computing. 1–4 (2008).
  11. Zhang Y., Bao Y., Zhao S., Chen J., Tang J.  Identifying Node Importance by Combining Betweenness Centrality and Katz Centrality.  2015 International Conference on Cloud Computing and Big Data (CCBD).  354–357 (2016).
  12. Ullah A., Wang B., Sheng J. F., Long J., Khan N., Sun Z. J.  Identification of nodes influence based on global structure model in complex networks.  Scientific Reports.  11 (1), 6173 (2021).
  13. Zhu J.-C., Wang L.-W.  An extended improved global structure model for influential node identification in complex networks.  Chinese Physics B.  31 (6), 068904 (2022).
  14. Mukhtar M. F., Abas Z. A., Baharuddin A. S. et al.  Integrating local and global information to identify influential nodes in complex networks.  Scientific Reports.  13 (1), 11411 (2023).
  15. Rossi R. A., Ahmed N. K.  The network data repository with interactive graph analytics and visualization.  Proceedings of the AAAI Conference on Artificial Intelligence.  29 (1), 4292–4293 (2015).
  16. Rücker G.  Network meta-analysis, electrical networks and graph theory.  Res Synth Methods.  3 (4), 312–324 (2012).
  17. Walker D. A.  JMASM9: Converting Kendall's Tau for Correlational or Meta-Analytic Analyses.  Journal of Modern Applied Statistical Methods.  2 (2), 525–530 (2003).
  18. Kendall M. G.  A New Measure of Rank Correlation.  Biometrika.  30 (1/2), 81–93 (1938).