Unified underpinning of human mobility in the real world and cyberspace

Yi-Ming Zhao,Wen-Xu Wang,Ying-Cheng Lai,An Zeng,Xiao-Yong Yan

doi:10.1088/1367-2630/18/5/053025

Abstract

Human movements in the real world and in cyberspace affect not only dynamical processes such as epidemic spreading and information diffusion but also social and economical activities such as urban planning and personalized recommendation in online shopping. Despite recent efforts in characterizing and modeling human behaviors in both the real and cyber worlds, the fundamental dynamics underlying human mobility have not been well understood. We develop a minimal, memory-based random walk model in limited space for reproducing, with a single parameter, the key statistical behaviors characterizing human movements in both cases. The model is validated using relatively big data from mobile phone and online commerce, suggesting memory-based random walk dynamics as the unified underpinning for human mobility, regardless of whether it occurs in the real world or in cyberspace.

Highlights

Human mobility in the real world and cyberspace plays an ever increasing role in the modern society and economy
We develop a random walk model with a single parameter to reproduce the statistical scaling behaviors of the three quantities characterizing human mobility
We demonstrate that, when this mechanism is incorporated into a standard random walk process, the analytically predicted behaviors agree, at a detailed and quantitative level, with those from two representative real data sets, one for real world and another for cyberspace movements

Summary

Introduction

Human mobility in the real world and cyberspace plays an ever increasing role in the modern society and economy. Modern research on human mobility dynamics began with the trajectory-based approach [9], e.g., by tracing the trajectories of dollar bills in the real world, which revealed a number of scaling relations such as a truncated power law in the distribution of the traveling distance. Analysis of mobile phone data demonstrated that individual travel patterns can be characterized by a spatial probability distribution, indicating the existence of universal patterns in human trajectories [10]. Human mobility in cyberspace and its relation to that in the real world were studied using big data analysis and phenomenological modeling [12]

Objectives

Methods

Results

Conclusion