Hierarchical core-periphery structure in networks

Area of Science:

• Network Science
• Computational Social Science

Background:

• Core-periphery structure is a fundamental concept in network analysis.
• Existing models often struggle to capture the diversity of network architectures.
• Understanding these structures is crucial for various fields, including sociology and computer science.

Purpose of the Study:

• To develop and apply a flexible network model for inferring core-periphery structures.
• To identify and differentiate between various types of core-periphery organizations in real-world networks.
• To explore hybrid structures combining core-periphery and community features.

Main Methods:

• Development of a flexible network model accommodating nested, hierarchical, and non-nested structures.
• Implementation of an efficient Monte Carlo scheme for model fitting to empirical data.
• Analysis of diverse real-world network datasets.

Main Results:

• Empirical distinction observed between traditional core-periphery (dense core, sparse periphery) and alternative structures (strongly connected core).
• Networks exhibit varying degrees of fit to these distinct structural types.
• Identification of hybrid structures merging core-periphery and community features, with non-overlapping cores resembling communities.

Conclusions:

• The proposed flexible network model effectively captures diverse core-periphery structures.
• Real-world networks display a spectrum of core-periphery organizations, including novel hybrid forms.
• The study provides computational tools for network structure analysis.