Research On Opinion Dynamics Modeling In Social Networks

In the information age,public opinion events have exploded,and public opinion communication has shown a variety of complicated ways,which has gradually become an important topic of complex scientific research.From the national level,mastering the initiative of public opinion will master the right to speak.The brewing and fermentation of public opinion will undoubtedly have a great impact on social development.It is imperative to scientifically master the law of public opinion development and monitor and guide it.Opinion dynamics studies the dynamic evolution process of opinion formation,restoring the specific characteristics of opinion interaction in the real world by defining micro-individual interaction rules,and recording the long-term evolution of opinion,during which consensus,polarization,splitting and other opinion dynamic phenomena may emerge.In order to better reveal the deep law characteristics of the phenomenon of public opinion,public opinion guidance,early warning and control and exploration of human world group behavior has an important role in promoting.How to better simulate the real world problems,simulation research and interpretation of the real world has become an important topic in the field of opinion dynamics.At present,opinion dynamics has made great progress and has been widely used in simulating real world simulation and public opinion prediction.Based on previous studies,this paper has done the following main work:（1）A simulation model of information cocoon based on opinion dynamics is proposed.At present,the research on information cocoon still stays at the level of qualitative research.Qualitative analysis is difficult to simulate the interaction background in the real world.It is not intuitive to analyze the formation mechanism of information cocoon,and it is not profound and specific to grasp the comprehensiveness and effect of influencing factors.In order to explore the internal mechanism of the information cocoon,based on the theory of system science,this paper modeled the information cocoon from the perspective of opinion dynamics and used Netlogo simulation analysis.Based on the community structure,the model of the information cocoon was constructed,in which the trust threshold was introduced so that different communities could interact with each other according to a certain threshold.The difference of trust threshold between inside and outside the community is used to express the difference of values between different communities.The simulation results show that:information cocoon exists and has deep and shallow levels;Increasing the capacity of communication with heterogeneous community has a positive effect on restraining cocoon.Establishing a mechanism to increase the connection between heterogeneous communities is helpful to inhibit the occurrence of cocoon chamber phenomenon.In the final state,the opinions tend to be close to the mainstream opinions,and the corresponding measures to break the "cocoon" are put forward.（2）A opinion dynamics simulation model based on environmental factors is proposed.Environmental factors have a profound impact on the update of views,and the extent of the impact of environmental factors is not well understood.Community structure in the real world of social networks,to explore the effects of environmental factors on community structure,in this paper,based on the theory of system science thought,from the point of opinion the dynamics modeling and using Netlogo simulation analysis to its,the introduction of information entropy,the environmental factors into the entropy value,and establish the influence function affect the opinion updates,Based on this,a opinion dynamic model of the influence of environmental factors is constructed.The simulation results show that the influence of environmental factors on community structure is weaker than that on scale-free network,because the internal characteristics of community structure make the community have a strong tendency of consistency,and it is difficult to change the opinions.However,the structure of scale-free network is relatively loose,and environmental factors have a greater effect on it,and the opinions are easy to collapse and change.When the degree of mixing is low,that is,when a certain point of opinion tends to have obvious advantages,its influence is great and the point of opinion can converge to the final state faster.At the same time,the model also puts forward some suggestions for breaking the information cocoon.（3）The opinion dynamics model of community structure based on fuzzy similarity is proposed.Real world opinion is fuzzy,in order to study the evolution process of fuzzy perspective community point of view,this article is based on DW model community opinions,and introduce the fuzzy similarity mechanism to quantify the degree of similarity between two interactive opinion,thereby improving opinion update rules,in order to build a community opinion evolution model based on fuzzy similarity.The emergence of ideas under information cocoon and fuzzy similarity model is compared.The simulation results show that when the trust threshold of similarity is less than 0.5,the opinions always converge to a opinion cluster,and the interaction conditions of opinions under fuzzy similarity model are loose and convergence is faster.Increasing the number of connections between communities speeds up the updating of ideas.The distribution of initial opinions has a certain effect on the final state of opinions.This model verifies the existence of information cocoon in a more realistic way.