Modeling the Heterogeneity of Human Dynamics Based on the Measurements of Influential Users in Sina Microblog

Chenxu Wang,Xiaohong Guan,Tao Qin,Tao Yang

DOI: https://doi.org/10.1016/j.physa.2015.02.024

2015-01-01

Abstract:Online social network has become an indispensable communication tool in the information age. The development of microblog also provides us a great opportunity to study human dynamics that play a crucial role in the design of efficient communication systems. In this paper we study the characteristics of the tweeting behavior based on the data collected from Sina Microblog. The user activity level is measured to characterize how often a user posts a tweet. We find that the user activity level follows a bimodal distribution. That is, the microblog users tend to be either active or inactive. The inter-tweeting time distribution is then measured at both the aggregate and individual levels. We find that the inter-tweeting time follows a piecewise power law distribution of two tails. Furthermore, the exponents of the two tails have different correlations with the user activity level. These findings demonstrate that the dynamics of the tweeting behavior are heterogeneous in different time scales. We then develop a dynamic model co-driven by the memory and the interest mechanism to characterize the heterogeneity. The numerical simulations validate the model and verify that the short time interval tweeting behavior is driven by the memory mechanism while the long time interval behavior by the interest mechanism.