Rumor Spreading On Complex Networks Based On Dynamics Of Human Behavior

In this paper, we study the impact of the time-activity patterns of human behaviors on rumor spreading in complex networks and the impact of the space-activity patterns of human behaviors on the density distribution of people in a limited area.Rumor spreading is a typical social phenomenon in our daily life. The harmfulness in crisis or emergency is considerable. Therefore, rumor spreading is always an important topic in the study of complex networks. Firstly, we study the impact of the time-activity patterns of human heavy-tailed non-Poisson behaviors on rumor spreading in complex networks. Compared with the classical spreading network which assumes people mixed evenly, complex network is closer to the real social network. We simulate the scale-free network as the social network. Nowadays, as the development of computer, more and more evidences reveal that the inter event times of human behaviors are of high heterogeneity and its distribution P(τ) possesses a heavy tail. The individual time-activity patterns will affect the effectiveness of rumor spreading in the population. Researchers introduced the SIR model to focus on the range and efficiency of rumor spreading. Compared with the Poisson pattern case, we show that the time-activity patterns of the heavy-tailed non-Poisson human behaviors only slow down the spreading significantly, while the number of the affected individuals does not change.Furthermore, we believe people spread the rumor with "timeliness" in our real life, the infective individual would not spread the rumor forever. After the time span τρ, the infective one turns into the recovered state automatically which never spread the rumor. Therefore, we define τρ as the time limit that the information stays popular, or equivalently, the time span that the infective agents keep interest in the information. we find that the heavy-tailed human behaviors will hinder the spreading for large τρ. However, heavy-tailed human behaviors can promote the rumor spreading for small τρ. These results can help us understand the rule of rumor spreading, control and limit the outbreak of rumors.In recent years, researchers have focused not only on the time pattern of human behavior, but also extended to the study of the spatial characteristics of human behavior. A lot of hot issues researches need to consider the density distribution of people in a limited area, such as emergency evacuation, information dissemination, the research of epidemics and so on. We collect the spatial density of people in Tian’anmen Square and study its statistical characteristics.We find that the spatial density of people follows a power-law distribution. This reveals that individuals are scattered in most places, meanwhile people can also be highly concentrated in a few places. Moreover, we propose an attracted walk model, the random walks are biased because of the attraction from the surrounding points, which results in highly inhomogeneous of the spatial density of people. The model matches our empirical data well.In practice, compared with the uniform one, SDP power-law distribution will greatly affect the speed and scope of the information or disease spread. It is difficult for virus or information to spread in SDP sparse areas, even already dead before spreading out; however, information or virus will spread to a large area in a short time in densely SDP. On the other hand, we must also consider power-law distribution of SDP when we design of emergency evacuation passage in a limited area. It should be added emergency channel in SDP crowded places, meanwhile reduced the channel in SDP scattered places. This will not only reduce construction costs, more importantly, can optimize evacuation time, fewer casualties and property losses. Our research provides a theoretical basis for emergency management, information spreading, and the related fields.