Motif formation and industry specific topologies in the Japanese business firm network. (18th May 2017)

• Record Type: Journal Article
• Authors: Maluck, Julian; Donner, Reik V; Takayasu, Hideki; Takayasu, Misako
• Is Part Of: Journal of statistical mechanics. (2017:May)
• Publication Date: 2017-05-18
• Subjects: 11 -- 12; Statistical mechanics -- Periodicals; Mechanics -- Statistical methods -- Periodicals; 530.1305
• DOI: 10.1088/1742-5468/aa6ddb
• Languages: English
• ISSNs: 1742-5468

Abstract

Abstract: Motifs and roles are basic quantities for the characterization of interactions among 3-node subsets in complex networks. In this work, we investigate how the distribution of 3-node motifs can be influenced by modifying the rules of an evolving network model while keeping the statistics of simpler network characteristics, such as the link density and the degree distribution, invariant. We exemplify this problem for the special case of the Japanese Business Firm Network, where a well-studied and relatively simple yet realistic evolving network model is available, and compare the resulting motif distribution in the real-world and simulated networks. To better approximate the motif distribution of the real-world network in the model, we introduce both subgraph dependent and global additional rules. We find that a specific rule that allows only for the merging process between nodes with similar link directionality patterns reduces the observed excess of densely connected motifs with bidirectional links. Our study improves the mechanistic understanding of motif formation in evolving network models to better describe the characteristic features of real-world networks with a scale-free topology.