Research On Group Event Evolution And Risk Identification

Period of social transformation, social contradictions increasingly fierce, various socialproblems highlight. In recent years, the frequent mass incidents have become a prominentmanifestation of social contradictions in the period of social transformation in our country,seriously disturb the social order, affect social stability and development. Therefore, it isextremely important to study the evolution rules,risk discrimination and risk response of massincidents.The first chapter of this paper provides a brief overview of the results of previous studies anddeficiencies.The second chapter was studied for the dynamic evolution process of mass incidents.Considering the mass incidents as a dominant social conflict phenomenon, has the gamecharacteristics, the study decides to adopt the forefront of modern economics theory-evolutionarygame theory as the theory tool to establish the evolutionary game model.Within the Perfecttheory frame, the replicated dynamic equation has been proposed and evaluated in order to showthe dynamic development process of mass incidents, at the same time, based on modelparameter analysis, the complex interest relationship of mass incidents of game participants andits influence on the development of the incidents evolution has been revealed completely whichdo benefit to explain analyze the evolutionary rules of mass incidents and propose correspondingstrategies for risk emergency management.Facing to the shortage of the quantitative study of mass incidents risk identification inacademic circles now, the third chapter of this paper carried out a new appraising methodwhich is based on the BP neural network. The method was used on the basis of the previousresearch on appraising method of mass incidents risk which already exist. Research gives afeasible evaluation procedure through the establishment of mass incidents risk evaluation indexsystem and the BP neural network risk assessment model. On reckoning, the paper used theMATLAB neural networks tool–box (NNT) to design the net and calculate. Training and testingby studying a sample data, the results shows that the accuracy of risk assessment model is veryhigh, the training accuracy can reach10-9magnitude, the output precision can reach10-5magnitude, and a real mass incident has already occurred as an example to demonstrate the modelhas a good generalization performance, theoretical accuracy and assess effectiveness.Finally, combining with the research results of the second and third chapter of this paper, thefourth chapter proposed some risk prevention strategies on the basis of the analysis of theingredients of a game, meanwhile, the forewarning management system of mass incidents is setup and the theoretical design of its operation process is provided, what’s more, some effectiveemergency disposal methods are given.In short, the studies hope to provide an effective andaccurate model for risk identification and offer some scientific suggestions for dealing with massincidents, because these are of great practical significance to ensure smooth running of the social,reduce the loss of society, maintain social stability and build a harmonious society.