| With the rapid development of technology, information spreading evolved from the first word of mouth to the users’communication based on the Internet, which made the speed and the breadth of information spreading enhanced significantly. The characteristics of the information spreading have been caused by the change in the way of spreading information, and they are very different from the past. Now most models used in the study of information spreading are often derive from SIR model of disease spreading, or the variation based on this model. But there are some essential differences between the two phenomena:human subjective initiative plays an important role in the information spreading, but people are passive completely in the disease spreading. Recently, some modeling works which are based on the characteristics of current information spreading have been done, but this kind of work is few and coarse relatively, so it is one of the hot spot of current information spreading research to study the characteristics of information spreading, and to explore spreading model which is more in line with the actual, even to model on specific dispersal unit.In this paper, the general rule of information spreading, the relationship between the scope of secondary spreading and the distribution of eventual spreading scale, as well as the average spreading scale will be studied, the model which is more in line with the actual life will also be established, the details are as follows:(1) Studying the general rule of information spreading:the distribution of eventual spreading scale, the average spreading scale, and the probability density distribution of eventual spreading scale. A spreading model (ZLZ model) which can well explains Dr.Centola’s experiments on behavior spreading in online social network has been put forward by Zheng Muhua.etc, and this model can also be applied to describe the spreading of information, mainly for the information whose credibility is not strong. However, the SIR model tends to describe the spreading of information whose credibility is strong. Therefore, the two models will be used by us to study the general rule of information spreading. It’s found that the distribution of eventual spreading scale is not concentrate in area near the average spreading scale, eventual spreading scale appears dragon king:Power law distribution and tail-raising, the wide-ranging spreading probability is very small.(2) Studying the relationship between secondary spreading range and the distribution of eventual spreading scale, and the average spreading scale, also studying the probability density distribution of eventual spreading scale under the different secondary spreading range. The secondary spreading range been considered, and the spreading model which is put forward by Zheng Muhua etc. and SIR model been applied in our work. It’s found that the distribution of eventual spreading scale on the different network structure is obviously different. The greater the secondary spreading range is, the greater the average spreading scale will be, but when the secondary spreading range reaches a certain size, the average spreading scale is not sensitive to the secondary spreading range. Secondary wide-ranging spreading probability is very small.(3) Considering individual difference, and put forward a new model of information spreading. The work which works by Zheng Muhua etc. been analysis and researched by us, it’s found that theirs simulation results are only accord well with Centola’s experimental results in qualitative, there are some gaps in the quantitative. For everyone has theirs unique properties in view of the real life, the individual difference will be considered and we make it meet the Poisson distribution, further their model will be improved. It’s found that the simulation results can be accord with the experimental value under certain parameter, it’s also found that the initial spreading probability impacts spreading stronger than social reinforcement. |
PDF Full Text Request