Font Size: a A A

Research On Modeling And Simulation Of Emergency Network Public Opinion Propagation Based On MAS

Posted on:2016-08-21Degree:MasterType:Thesis
Country:ChinaCandidate:X ZhouFull Text:PDF
GTID:2348330536967361Subject:Control Science and Engineering
Abstract/Summary:PDF Full Text Request
With the swift development of the Internet,the network has deeply integrated into people's daily lives and has become an important part of society.The network applications possessing capabilities of information publishing,diffusion,and interaction and searching,such as micro-blog,blog,QQ,micro letters and forums,have become the major sources and amplifiers of public opinion.Currently,public opinion on the internet maintains a rapidly growing momentum,which cannot be ignored because it is a barometer of social mood.The public opinion on the Internet generated by emergencies often reflects the public attitude towards sensitive events,which may lead to a seriously social crisis with improper handling.Therefore,it is necessary to do some researches about network public opinion.In this condition,it is an effective method to study the emergency propagation process of network public opinion from the perspective of macro and micro.At the macro level,this paper proposes a "Four Phases Five Points" model of emergency network public opinion based on life cycle theory of information dissemination.While at the micro level,the paper proposes a behavior model which called "Psychology-Behavior" processing information model of netizens based on Multi-Agent Systems theory.The following parts are major works of research.(1)This paper proposed a new model about the evolution of emergency network public opinion propagation which is called "Four Phases Five Points" model.In previous work of the evolution process of propagation on emergency network public opinion,most models are qualitative.Drawing the lessons of previous researches,this paper proposes a quantitative model of four phases.Firstly,according to the propagation characteristics of the network public opinion,five important feature points of emergencies network public opinion can be extracted,which is generation point,outbreak point,the peak point,regression point and the dying point.Secondly,according to the five feature points,public opinion can be divided into four stages,namely,generation phase,diffusion increasing phase,diffusion descent phase and dying phase.Thirdly,in order to analyze quantitative research,several quantitative indicators should by set for each stage.Finally,due to ?Four Phases Five Points? model,the paper does some researches and analysis about micro blog data on emergency in 2014,and some meaningful conclusions are put forward.(2)Based on Multi-Agent System,this paper presents “Psychology-Behavioral " processing information model of netizens.On the basis of previous results,we propose a new model of information processing behaviors about netizens.Netizens are divided into two types,which are called event-related netizens and event-unrelated netizens separately.Based on media literacy theory,this article simplifies the information processing behaviors of netizens into three stages: information receiving stage,information cognitive stage,and information dissemination stage.Information received stage indicates whether netizens have the ability to receive the information.The information cognitive stage shows how netizens process information,while information dissemination stage represents whom the netizens would propagate the information to.In this paper,the information cognitive stage is highlighted because it is the hardest part of netizens.As for event-related netizens,the information cognitive stage is composed of two models: the judgment behavior model based on psychological feature and opinion decision model based on Markov Chain.While as for event-unrelated netizens,the information cognitive stage can be simplified by opinion dynamic model.Finally,we construct public opinion simulation system according to ?Psychology-Behavioral" model.Public opinion simulation system consists of information,network between netizens,and netizens.By learning the regulation of micro blog data of emergency in 2014,the parameters are assigned the initial values.After that,the public opinion simulation system simulates the "Tianjin Port explosion" event.In general,the results show that this model has a certain scope,and the network evolution trend of public opinion of emergencies can be predicted to some extent.
Keywords/Search Tags:Network Public Opinion of Emergency, Theory of Life Cycle, Multi-Agent System, Behavior of Processing Information of Netizen, Public Opinion Simulation
PDF Full Text Request
Related items