Study On Information Diffusion Via Social Network Based On Biological Intelligence

With the rapid development and extensively application of network technology,the online social network has become an essential way to exchange information as a substitute for the traditional media.As the main body of the social network,users are not only the disseminators but also the receivers for information.Utilizing online social network services,users share their views,and communicate with other users who have similar habits and interests,consequently,build their social relationships.Since higher autonomy and self-organization of each user in the social network,studying on the characteristics of individual behavior and information diffusion becomes the hottest topics in the field of interdisciplinary research.In addition,control the direction of the information spreading,minimize the cost,and choose the appropriate nodes are also the great challenges for modeling the social network.The main contribution of the thesis lies in:1)Extract the data of hot topics from micro-blog are extracted,and clean and format these data.Then the distribution and correlation of several key features of the users are analyzed.Based on the characteristics of the users,the information diffusion model is proposed.Finally,it has been verified that the topic developing trend can be well predicted by the proposed model in MicroBlog,which matches well with the real situation.2)Through the analysis of the immune response mechanism in organisms,we found that the process of information diffusion has great similarities to the immune system which can keep the homeostasis stable by responding the intrusive antigens.Inspired by the immune system,the model is established by considering the social reinforcement.Experimental analysis prove that the model can restore the process of the information spread very well.3)Inspired by the immune endocrine system characteristics,especially mechanism of hormone transmission and regulation,an endocrine-immune-based information diffusion(EIID)model is established to solve the controllability of the information diffusion.Experimental results show that the proposed EIID model can control information diffusion and achieve the desired propagation target quickly.4)Firstly,according to the characteristics of the social network,an intelligent service model of social network is set up.Then inspired by neuroendocrine system,especially the regulation process of the stress response,a model of intelligent information service is set up.The experimental results show that the model can respond quickly to the service requests generated in the network.Finally,a conclusion is made for the whole contents of this dissertation,in which both the disadvantages and the perspectives of this field for the next step have also been discussed.