Table 5.

Summarizes social network applications, algorithms, and network metrics.

Reference	Application	Feature	Algorithm	Network Measure
[34,47]	To detect criminal networks	Call features	Girvan–Newman and Fruchterman–Reingold	Degree centrality, eigenvector centrality, closeness centrality, transitivity, betweenness centrality, and transitivity
[3,35]	To detect criminal networks	Call features	Concept space approach and Prim’s algorithm	Vertex-centric, edge-centric
[84]	To detect customers who are likely to fail to pay their mobile bill	Call feature and spatiotemporal features	SLPA	Closeness centrality and reciprocity measures
[85]	To detect users’ social interactions	Call feature	Bron and Kerbosch’s	Persistence, disparity, and reciprocity measures
[86]	To detect ethnic communities in Ivory Coast	Call and spatiotemporal features	Louvain	Asymmetries and assortativity coefficient
[87]	To detect human spatial interactions in China	Spatiotemporal features	Infomap, Louvain, and REDCAP	Degree, strength, rich-club coefficient, and assortativity coefficient
[88]	To detect socio-economic groups in Ivory Coast	Call and spatiotemporal features	Louvain	Rich-club coefficient and PageRank
[26]	To detect he spatial interactions of communities in Milan, Italy	Spatiotemporal features	Louvain	Betweenness centrality, degree centrality, and PageRank
[89]	To detect individual ‘s spending behaviors	Spatiotemporal features	Louvain	Diversity, radius of gyration, and homophily
[90]	To detect socio-economic communities in Santiago, Chile	Spatiotemporal features	Louvain	Segregation measures (i.e., isolation metric)
[91]	To detect urban communities in Dublin, Ireland	Spatiotemporal features	Louvain	Newman’s modularity metric