Research On Mechanism Of Information Dissemination And Prediction Approaches In Social Network

With the rapid development of ideas and technologies of Web2.0,Social Network,such as Facebook,Twitter,LinkedIn and Sina Weibo,has changed the people's way of life,and it also plays an important role in the people's social life,way of sharing information and discussion.With the constant development of social networks,the inner connection between social networks and real-world society is becoming closer,which has significant practical meanings and research value.As a result,Social Network has been one of the most popular research fields in recent years.First of all,whether we are able to obtain the information dissemination data quickly and accurately? whether the mechanism of traditional fields is still applicable in social networks?Secondly,whether the behavior of information dissemination in social networks can be predicted? whether there is an appropriate measurement and method to represent the information dissemination structure in social networks? whether the prediction of information dissemination in social networks can be modeled? These are all becoming open and challenging issues.At last but not least,how to integrate these methods into a practical application system which needs further validation.As a result,based on the above research background and relevant issues,the main investigation of this paper includes:1.Research on the mechanism of information dissemination in social networks: Firstly we analyze the impact of social network structure on the mechanism of information dissemination,and we propose an efficient algorithm which can generate the information dissemination tree in social networks.Then,based on the real-world information dissemination datasets in Sina Weibo,we find that the Spiral of Silence Theory is still applicable in social networks through a large amount of experiments.2.Research on the prediction of information dissemination in social networks: Firstly,we introduce the basic idea for prediction of information dissemination structure.Secondly,we determine the relationship between the information dissemination structure and modularity through a real case.After that,we propose an information dissemination structure prediction model which is based on the triple exponential smoothing model.We finally test and evaluate the proposed prediction model.3.Research on the social networks information dissemination analysis and prediction system: Based on the above research results,we design and implement the SNIDAPS system,we describe the structure,workflow,and key modules of SNIDAPS system in detail.Besides,the experimental results appear that SNIDAPS system can availably and efficiently analyze and predict information dissemination in social networks.This paper combines research methods and theories of many research fields to investigate the problem of information dissemination mechanism and prediction in social networks,including Communication,Data Mining,and Complex Network and so on.Through this paper,not only can we have a better understanding of information dissemination in social networks,but also can effectively predict the information dissemination structure.In addition,these research results also can be used in advertisement,personal recommendation and social management.