Evolution And Experimental Analysis Of Consumer Belief In Social Network Based On Agents

The purpose of the evolution modeling of consumer belief in social network is to study the changes of consumer belief in the process of social interaction.Research on the evolution of consumer beliefs in social networks has theoretical significance and practical value.However,the complexity of the social network structure,the different signal structures of consumers,and the multi-factors that affect the changes in consumer beliefs have brought difficulties to this work.In recent years,scholars have vigorously studied the dynamics of complex network concepts,which provides the possibility for the study of the evolution of individual consumer beliefs in this paper.This thesis adopts the agent-based method and simulates social interactions on social network models,and studies the evolution of individual consumer beliefs from a micro perspective.Firstly,technical requirements are extracted from the typical user consumption scenarios in social networks.Then,according to the widespread similarity of social networks and the influence of network structure,a social network model based on similarity is proposed.The similarity is mainly considered from the internal attributes and external behavior activities of the nodes,and the influence of the network structure mainly considers the neighbor structure and group structure,and the experimental analysis is carried out.Moreover,three different consumer signal structures are defined,and a consumer belief update model based on feedback cyclic strategy learning is proposed.Based on the framework of information diffusion-information aggregation-cognitive update,the model analyzes the influence of consumers' different signal structures on information diffusion,and the aggregation of strategic learning information without event intervention and event influence,then,it analyzes the updating of consumer creed in cognitive system.Finally,the agent-based method is used to analyze the change of users' consumption beliefs in social interactions.Experiment results show that proposed models can reflect the reality well.Agents with different signal structures have different initial consumption creeds and different rates of change,but no matter what value the agent's consumption belief starts from,Finally,it will gradually converge to a certain stagnation point or a certain horizontal line,the result reflects the effectiveness and rationality of the proposed strategy learning feedback loop model.