Mining: communities: Community detection algorithms

Description

Various algorithms to compute communities in multiplex networks, based on frequent itemset mining (abacus), adjacent cliques (clique percolation), modularity optimization (generalized louvain) and random walks (lart). get_community_list_ml is a commodity function translating the result of these algorithms into a list of node identifiers, and is internally used by the plotting function.

Usage

abacus_ml(mlnetwork, min.actors=3, min.layers=1)
clique_percolation_ml(mlnetwork, k=3, m=1)
glouvain_ml(mlnetwork, gamma=1, omega=1, limit=0)
infomap_ml(mlnetwork, overlapping=FALSE, directed=FALSE, self.links=TRUE)
modularity_ml(mlnetwork, comm.struct, gamma=1, omega=1)
get_community_list_ml(comm.struct, mlnetwork)

Arguments

mlnetwork

A multilayer network.

min.actors

Minimum number of actors to form a community.

min.layers

Minimum number of times two actors must be in the same single-layer community to be considered in the same multi-layer community.

Minimum number of actors in a clique_ Must be at least 3.

Minimum number of common layers in a clique_

gamma

Resolution parameter for modularity in the generalized louvain and lart methods.

omega

Inter-layer weight parameter in the generalized louvain method.

limit

Limit parameter in the generalized louvain method: if the number of nodes exceeds this limit, then the modularity is computed on the fly without keeping the full data in memory.

overlapping

Specifies if overlapping clusters can be returned.

directed

Specifies whether the edges should be considered as directed.

self.links

Specifies whether self links should be considered or not.

comm.struct

The result of a community detection method.

Value

All community detection algorithms return a data frame where each row contains actor name, layer name and community identifier.

get_community_list_ml transforms the output of a community detection function into a list by grouping all the nodes having the same community identifier and the same layer.

References

Berlingerio, Michele, Pinelli, Fabio, and Calabrese, Francesco (2013). ABACUS: frequent pAttern mining-BAsed Community discovery in mUltidimensional networkS. Data Mining and Knowledge Discovery, 27(3), 294-320. (for abacus_ml()) Afsarmanesh, Nazanin, and Magnani, Matteo (2018). Partial and overlapping community detection in multiplex social networks. Social informatics (for clique_percolation_ml()) Mucha, Peter J., Richardson, Thomas, Macon, Kevin, Porter, Mason A., and Onnela, Jukka-Pekka (2010). Community structure in time-dependent, multiscale, and multiplex networks. Science (New York, N.Y.), 328(5980), 876-8. Data Analysis, Statistics and Probability; Physics and Society. (for glouvain_ml()) De Domenico, M., Lancichinetti, A., Arenas, A., and Rosvall, M. (2015) Identifying Modular Flows on Multilayer Networks Reveals Highly Overlapping Organization in Interconnected Systems. PHYSICAL REVIEW X 5, 011027 (for infomap_ml())

Examples

Run this code

# NOT RUN {
net <- ml_florentine()
abacus_ml(net)
clique_percolation_ml(net)
glouvain_ml(net)
infomap_ml(net)
# }