Shannon Entropy in Time-Varying Clique Networks

摘要

Recent works have used information theory in complex networks. Studies often discuss entropy in the degree distributions of a network. However, there is no specific work for entropy in clique networks. In this regard, this work proposes a method to calculate clique network entropy, as well as its theoretical maximum and minimum values. The entropies are calculated for the dataset of the semantic networks of titles of scientific papers from the journals Nature and Science for approximately a decade. Journals are modeled as time-varying graphs and each system is analyzed from a time sliding window. The results show the entropy values of vertices and edges in each window arranged in time series, and also suggest the moment which has more or less vocabulary diversification when this diversity turns the studied journals closer or move them away. For that matter, this report contributes to the studies on clique networks and the diffusion of human knowledge in journals of high scientific impact.