In the years, the researches have posed their attention on the study of networks analysing the nature, the composition, the relationship developed inside them, their construction and growth. Recently, particular attention has been posed on the analysis of complex networks and on the community detection analysis. The complex network has been defined as a network […] open, value-laden, directed, multilevel, multicomponent, reconfigurable systems of systems, and placed within unstable and changing environments (Boccaletti et al., 2014, p. 6). In this complexity, particular interesting is the identification of communities inside the network. It is possible to define the communities as subgroups of nodes with a density of internal connections larger than the density of external links. The aim of the community detection analysis is to identify the community structure inside the network in order to define the modular decomposition of the network. In literature many community detection algorithms are identified. To evaluate the capacity of these algorithms to identify the community, the algorithms have been applied on two different networks: the Zachary’s karate club network and Friendship network of a UK university faculty. The first is an indirect graph, as it is characterised to have edges that are not directed. On the contrary, the second is a direct graph, as it is composed by directed edges. The first findings evidence how each algorithm identifies specific communities inside the networks. These communities are composed in the most of cases by different nodes. Only in few cases, similarities have been detected. Moreover, it has been identified some problematic in the analysis of direct graph.

Four community detection algorithms for direct and indirect graphs

Contu, Giulia;Frigau, Luca;Romano, Maurizio
2019-01-01

Abstract

In the years, the researches have posed their attention on the study of networks analysing the nature, the composition, the relationship developed inside them, their construction and growth. Recently, particular attention has been posed on the analysis of complex networks and on the community detection analysis. The complex network has been defined as a network […] open, value-laden, directed, multilevel, multicomponent, reconfigurable systems of systems, and placed within unstable and changing environments (Boccaletti et al., 2014, p. 6). In this complexity, particular interesting is the identification of communities inside the network. It is possible to define the communities as subgroups of nodes with a density of internal connections larger than the density of external links. The aim of the community detection analysis is to identify the community structure inside the network in order to define the modular decomposition of the network. In literature many community detection algorithms are identified. To evaluate the capacity of these algorithms to identify the community, the algorithms have been applied on two different networks: the Zachary’s karate club network and Friendship network of a UK university faculty. The first is an indirect graph, as it is characterised to have edges that are not directed. On the contrary, the second is a direct graph, as it is composed by directed edges. The first findings evidence how each algorithm identifies specific communities inside the networks. These communities are composed in the most of cases by different nodes. Only in few cases, similarities have been detected. Moreover, it has been identified some problematic in the analysis of direct graph.
2019
9788891798701
Network; complex network; community detection algorithms; Zackary networks
File in questo prodotto:
File Dimensione Formato  
frigau 251-262FrancoAngeli Open Access.pdf

accesso aperto

Tipologia: versione editoriale
Dimensione 1.5 MB
Formato Adobe PDF
1.5 MB Adobe PDF Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11584/283048
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
social impact