Motif-based community detection in heterogeneous multilayer networks.

Area of Science:

• Complex Networks Analysis
• Network Science
• Data Mining

Background:

• Multilayer networks are crucial in complex systems, featuring both intralayer and interlayer connections.
• Existing community detection methods often overlook the heterogeneity of nodes and edges in multilayer networks.
• Research has primarily focused on multiplex networks, neglecting heterogeneous multilayer networks with diverse node and edge semantics.

Purpose of the Study:

• To address the limitations of current methods for community detection in heterogeneous multilayer networks.
• To propose a novel algorithm for identifying community structures in complex, heterogeneous multilayer networks.
• To investigate the relationship between network motifs and community structures in these networks.

Main Methods:

• Definition of communities and motifs, including interlayer motifs, within multilayer networks.
• Development of a motif-based modularity measure tailored for heterogeneous multilayer networks.
• Community structure detection achieved by maximizing the proposed motif-based modularity.

Main Results:

• The motif-based modularity community detection algorithm demonstrates superior performance on synthetic networks compared to classical methods.
• Experimental results show a significant relationship between network motifs and detected communities.
• The algorithm's applicability is validated through successful implementation on an empirical network, confirming its real-world practicality.

Conclusions:

• The proposed motif-based modularity approach offers an effective solution for community detection in heterogeneous multilayer networks.
• This method enhances the understanding of complex network structures by incorporating interlayer heterogeneity.
• The study provides a valuable tool for analyzing heterogeneous information within multilayer network systems.