Allometric Scaling of Mutual Information in Complex Networks: A Conceptual Framework and Empirical Approach

Complexity and information theory are two very valuable but distinct fields of research, yet sharing the same roots. Here, we develop a complexity framework inspired by the allometric scaling laws of living biological systems in order to evaluate the structural features of networks. This is done by...

Full description

Bibliographic Details
Main Authors: Eduardo Viegas, Hayato Goto, Yuh Kobayashi, Misako Takayasu, Hideki Takayasu, Henrik Jeldtoft Jensen
Format: Article
Language:English
Published: MDPI AG 2020-02-01
Series:Entropy
Subjects:
Online Access:https://www.mdpi.com/1099-4300/22/2/206
Description
Summary:Complexity and information theory are two very valuable but distinct fields of research, yet sharing the same roots. Here, we develop a complexity framework inspired by the allometric scaling laws of living biological systems in order to evaluate the structural features of networks. This is done by aligning the fundamental building blocks of information theory (entropy and mutual information) with the core concepts in network science such as the preferential attachment and degree correlations. In doing so, we are able to articulate the meaning and significance of mutual information as a comparative analysis tool for network activity. When adapting and applying the framework to the specific context of the business ecosystem of Japanese firms, we are able to highlight the key structural differences and efficiency levels of the economic activities within each prefecture in Japan. Moreover, we propose a method to quantify the distance of an economic system to its efficient free market configuration by distinguishing and quantifying two particular types of mutual information, total and structural.
ISSN:1099-4300