Fractal Scaling

Wang Cheng-Jun edited this page Dec 19, 2016 · 1 revision

计算传播学是计算社会科学的重要分支。它主要关注人类传播行为的可计算性基础,以传播网络分析、传播文本挖掘、数据科学等为主要分析工具,(以非介入地方式)大规模地收集并分析人类传播行为数据,挖掘人类传播行为背后的模式和法则,分析模式背后的生成机制与基本原理,可以被广泛地应用于数据新闻和计算广告等场景,注重编程训练、数学建模、可计算思维。

Clone this wiki locally

Emergence of fractal scaling in complex networks

Zong-Wen Wei and Bing-Hong Wang Phys. Rev. E 94, 032309 – Published 14 September 2016

ABSTRACT

Some real-world networks are shown to be fractal or self-similar. It is widespread that such a phenomenon originates from the repulsion between hubs or disassortativity. Here we show that this common belief fails to capture the causality. Our key insight to address it is to pinpoint links critical to fractality. Those links with small edge betweenness centrality (BC) constitute a special architecture called fractal reference system, which gives birth to the fractal structure of those reported networks. In contrast, a small amount of links with high BC enable small-world effects, hiding the intrinsic fractality. With enough of such links removed, fractal scaling spontaneously arises from nonfractal networks. Our results provide a multiple-scale view on the structure and dynamics and place fractality as a generic organizing principle of complex networks on a firmer ground.

http://journals.aps.org/pre/abstract/10.1103/PhysRevE.94.032309