Skip to main content
Applied Network Science

Table 2 Generalized Conventional Normalized Mutual Information (Lutov et al. 2019) between the communities of our algorithm and the compared traditional community detection algorithms (higher = more similar)

From: Detecting and generating overlapping nested communities

 

NMIM

NMIC

NMIL

adjnoun

0.686

0.911

0.538

celegansneural

0.567

0.678

0.657

dolphins

0.655

0.821

0.676

*karate

0.645

0.699

0.533

netscience

0.941

0.953

0.949

power

0.963

0.974

0.927

dswomen

0.842

0.836

0.530

*families

0.879

0.894

0.724

les_miserables

0.576

0.698

0.617

  1. The subscripts \(\textrm{M}\), \(\textrm{C}\) and \(\textrm{L}\) correspond to the algorithms MOSES, CFinder and Linkcomm, respectively. Graphs marked with an asterisk (*) were directly mentioned and analyzed in the article
Back to article page