Effects of degree-biased transmission rate and nonlinear infectivity on rumor spreading in complex social networks

Abstract

We introduce a generalized rumor spreading model and analytically investigate the spreading of rumors on scale-free (SF) networks. In the standard rumor spreading model, each node has an infectivity equal to its degree, and connectivity is uniform across all links. To generalize this model, we introduce an infectivity function that determines the number of simultaneous contacts that a given node (individual) may establish with its connected neighbors and a connectivity strength function (CSF) for the direct link between two connected nodes. These lead to a degree-biased propagation of rumors. For nonlinear functions, this generalization is reflected in the infectivity's exponent α and the CSF's exponent β. We show that, by adjusting exponents α and β, the epidemic threshold can be controlled. This feature is absent in the standard rumor spreading model. In addition, we obtain a critical threshold. We show that the critical threshold for our generalized model is greater than that of the standard model on a finite SF network. Theoretically, we show that β=-1 leads to a maximum spreading of rumors, and computation results on different networks verify our theoretical prediction. Also, we show that a smaller α leads to a larger spreading of rumors. Our results are interesting since we obtain these results regardless of the network topology and configuration.

An effective rumor-containing strategy

“…Empirical studies that started about two decades ago fully indicate that most realistic OSNs are heterogeneous rather than homogeneous [20,21]. In the past decade, therefore, much efforts were focused on rumor spreading models based on complex networks [22][23][24][25][26][27][28][29][30][31][32][33][34][35][36][37]. However, it is uncertain whether these models accurately characterize actual rumor spreading processes, because the models are derived through a series of approximations and do not perfectly accommodate the spreading network [38][39][40].…”

Complex dynamic behavior of a rumor propagation model with spatial-temporal diffusion terms

“…(i) Improvements on the previous rumor propagation models. The deterministic epidemiclike rumor propagation models proposed for online social networks in the majority of existing literatures are mainly based on ODE [24,30,35,42,45], which deals only with collective social processes over time without considering space factors for mobile online social networks. Though some researchers recently have begun to study both temporal and spatial patterns of information diffusion in social media by PDE [31,32,51], they may ignore the spatial-temporal delay phenomenon in information transmission.…”

“…Thus, distinguishing the sensitivity of these parameters is significant for understanding the mechanism of rumor propagation in online social networks. However, in most of the previous literatures, scholars randomly selected a parameter and then varied the values of this parameter to give a numerical simulation about their proposed models [23,24,44,45]. In this work, by applying sensitivity analysis in mathematics, we study the relationship between the density of spreading users and the parameters in our proposed model.…”

Rumor-Propagation Model with Consideration of Refutation Mechanism in Homogeneous Social Networks

“…Many other scholars have studied rumor propagation by considering the topological properties of social networks [13][14][15][16]. Zanette [17,18] and Buzna et al [19] studied the dynamic process of rumor propagation and found that there was a critical rumor-propagation threshold that must be exceeded before a rumor can propagate in small-world networks.…”