Research On The Evolution Of Consumer Network In Decentralized E-commerce Environment

The concept of "e-commerce" was proposed as early as 1990 s by American scholar Kalakota.It has gone through three stages,the current e-commerce 3.0 stage.E-commerce platforms mainly recruit users in the network as opinion leaders,and realize value co-creation through UGC(user-generated content)and private traffic.In this process,opinion leaders use their own influence as the key node of merchants’ marketing,and realize marketing split through interactive behaviors of interpersonal relationships among users.To a certain extent,it shares the role of centralized resource redistribution of the platform,and this form of e-commerce is called "decentralized e-commerce model" in this paper.At the level of user viewpoint change,this paper considers that in the decentralized e-commerce network,a node is rationally and continuously influenced by its connected neighbor nodes,and the probability of viewpoint change of this node is also rational and continuous.Therefore,the Ising model in viewpoint dynamics is used in this paper as a research means to analyze the group decision making and viewpoint consistency evolution of this network.It aims to achieve accurate marketing of e-commerce platforms,thereby increasing the stickiness of platform users and occupying market share.However,the influence of the traditional Ising model neighbor node on the view of the target node is random and irrational,and the change probability of the view is random and discrete,which is inconsistent with the characteristics of decentralized e-commerce network.In order to improve the above shortcomings,this paper first introduces rational trust threshold and trust interval parameters to obtain an improved Ising model of continuous temperature factor function.After python simulation,it is found that when the rational trust threshold and trust interval parameters are larger in the decentralized e-commerce network,the views of users in the network will gradually converge.At the level of user purchasing behavior,this paper considers that users in the decentralized e-commerce consumer network can choose to enter a certain area of interest,exit from a certain area of interest,and rejoin the withdrawn area of interest due to changes in interest preferences.In addition,new users are more likely to pay attention to influential opinion leaders recommended by the system.Users in the network are also likely to enter multiple areas of interest at the same time.Therefore,this paper makes the following improvements on the traditional BA model and divides the evolution of the decentralized e-commerce consumer network into two stages: The first stage is the formation stage of the decentralized e-commerce consumer network based on the traditional BA model,and the second stage is the evolution stage after the formation of the decentralized e-commerce consumer network,which is used to describe the evolution process after the stability of the network,including the node edge fracture mechanism and the priority connection mechanism based on the node vision.At the same time,in the first stage,a community attribute list is set for each node in the network to store the community to which the node belongs.After the network evolution,the community attribute list of the node is updated in time to record the changes of the community to which the node belongs in the process of network evolution.In the second stage,the joint edge breaking mechanism and the priority connection mechanism based on the node vision are introduced.The simulation results show that the improved BA model has better evolutionary effect.Finally,personalized recommendation is made according to the improved model,so as to achieve the effect of precision marketing.