A repository to conduct a comparative analysis on community-aware centrality measures.
Following is the reference for the paper: Rajeh, S., Savonnet, M., Leclercq, E., & Cherifi, H. (2022). Comparative evaluation of community-aware centrality measures. Quality & Quantity, 1-30. It can be accessed at: https://arxiv.org/pdf/2205.06995.pdf
Comments and questions are welcome, please contact: stephany.rajeh(at)u-bourgogne.fr
The sources of datasets used in the study are available within the article.
Notebook: NetworkName-CentralityCalculation.ipynb: Calculates the community-aware centrality for a given network.
Notebook: NetworkName-CorrelationHeatmap.ipynb: Computes the correlation for all possible combinations between the community-aware centrality measures and represents the values in a heatmap.
Notebook: NetworkName-SIRSimulations.ipynb: Runs the SIR simulation of the EoN library (https://epidemicsonnetworks.readthedocs.io/en/latest/) using the community-aware centrality measures ranks plots the ΔR.
Those 3 notebooks are used for each of the 15 networks under study.
Notebook: CorrelationAndViolinPlotsAcrossNetworks.ipynb: Computing the correlation across and between networks and the violin plots of the correlation distribution.
Notebook: LinearRegression.ipynb: Running a linear regression to study the relationship between the correlation between the community-aware centrality measures and the community structure strength.