Mesoscopic structures reveal the network between the layers of multiplex data sets.

Area of Science:

• Complex systems science
• Network science
• Information theory

Background:

• Multiplex networks consist of nodes connected by various interactions across different layers.
• Analyzing these complex systems is vital for understanding their structure and solving inference problems.
• Existing methods struggle to fully characterize the mesoscopic structure of multiplex networks.

Purpose of the Study:

• To develop a novel information theory method for analyzing multiplex networks.
• To extract the underlying network structure connecting different layers of multiplex data.
• To characterize mesoscopic similarities and community structures within multiplex networks.

Main Methods:

• An indicator function based on network ensemble entropy was developed.
• Clustering techniques were applied to identify communities in the 'network of networks'.
• The method was applied to the American Physical Society (APS) Multiplex Collaboration Network.

Main Results:

• The study successfully extracted the network connecting layers in multiplex data.
• Mesoscopic similarities between network layers were quantified using entropy.
• Community structures within the 'network of networks' were identified.
• The analysis revealed interplay between collaboration networks and knowledge organization in physics.

Conclusions:

• The proposed information theory method effectively characterizes multiplex network structures.
• The approach provides insights into the organization of scientific knowledge and collaboration patterns.
• This method offers a powerful tool for analyzing complex multilayered systems.