Skip to main content

Table 7 (Disagreeable ideas) Modularity measures, detected communities, and significant communities at level 0.05 (based on the Wilcoxon rank-sum test)

From: Contextual polarity and influence mining in online social networks

Community detection algorithm Modularity Number of communities Significant communities
With weight  
 Fast Greedy 0.44 11 10
 Multi-level 0.44 9 9
 Edge Betweenness 0.41 15 9
 Leading Eigenvector 0.41 15 9
 Info Map 0.41 34 8
 Walk Trap 0.39 23 7
 Spin Glass 0.25 15 8
 Label Propagation 0.14 9 4
Without weight  
 Fast Greedy 0.47 8 8
 Multi-level 0.44 9 9
 Edge Betweenness 0.45 20 8
 Leading Eigenvector 0.43 20 8
 Info Map 0.42 35 7
 Walk Trap 0.38 32 6
 Spin Glass 0.25 11 8
 Label Propagation 0.05 9 4
  1. Each community structure is detected by one of the eight algorithms, with and without sentiment scores as weight