| China’s Internet industry has come into the era of Web 2.0 as various smart mobile terminal devices gain increasing popularity and diverse instant messaging software and platforms develop.Therefore,the network,compared with traditional platforms of expression of public opinions,has gradually got pushed into the foreground where the public mood and behavior tendency are reflected.With its openness,the network has become a public platform where both individuals and social organizations may participate in the administration and discussion of state affairs,declare their positions and express their opinions,a passageway through which information and public opinions may be conveyed rapidly,as well as a front with the clashes of a variety of social thoughts,different requests for interests,and multiple ideologies.In this connection,the analysis of and response to network public opinions have become another vital task in the network-based society today.However,a major challenge to current research on the network public opinion lies in that traditional data mining techniques are insufficient to deal with mass data with huge noise brought by information redundancy and the radical change of ways of information dissemination.In addition,the lack of subdivision of diverse network public opinions and the lack of different analysis models corresponding to those network public opinions with different characteristics are also major limitations.In the context of big data environment,this dissertation,based on a combination of qualitative and quantitative analyses,conduct a series of studies on such three typical types of network public opinions as Internet rumors,online public opinions of college students and emergent public health incidents in terms of the transmission mechanism,early warning and decision-making mechanism as well as evolution mechanism of the network public opinion.Also,the dissertation,based on model analysis,proposes management and coping strategies corresponding to different types of network public opinions.The present study,based on the kinetic theory of infectious diseases,establishes the Internet rumor spreading analysis and judgment models respectively with a saturated contact rate and a nonlinear contact rate,and makes a quantitative analysis of the Internet rumor based on the theories of balance and stability of the dynamical system.Findings are as follows.A threshold value R0 exists in the spreading of Internet rumors and an internal non-zero balance point will exist when R0>1,that is,rumors may break out on a large scale in the system if allowed to develop by themselves;increased netizens will not impact the spreading of Internet rumors when the number of netizens accord with the Logistic curve;measures to increase the psychological effect coefficient may help control the spreading the Internet rumors more efficiently due to the high sensitivity of the threshold value to the variation of that coefficient;revealing false information and its disseminators is far more efficient in management than positive publicity.In light of the spreading characteristics of public opinions regarding emergent public health incidents,the present study constructs the directed weighted dynamic network structural model with the introduction of the Deffault model.In addition,it conducts a simulated analysis of the established network public opinion evolution model by Matlab tools,and thus verifies the validity and rationality of its model and analyzes the major factors that impact the evolution and spreading of network public opinions.Research findings show that the directed weighted dynamic BBV network is scale-free,which conforms to the features of online social network structure.The results also show that the government’s attitude r and media attention λ may exert remarkable influence on network pubic opinions.Therefore,it may effectively help eliminate social panic and maintain stability if the government and major public media may,as authorities,timely reveal relevant information and intensify guidance.In terms of the judgment of early warning levels for network public opinions of college students,the present study establishes the qualitative and quantitative judgment model of network public opinions based on intuitionistic fuzzy reasoning and analytic hierarchy process(AHP).It proposes the judgment of network public opinion warning levels by intuitionistic fuzzy reasoning with the importance of the topic,public reaction,and relation between the pubic and the topic as its participation factors,so that the proximate intuitionistic fuzzy set will be the specific warning level after the degree of membership of each factor is calculated through the intuitionistic fuzzy comprehensive judgment.The present study also proposes the decomposition of the goal into multiple index levels through AHP and the determination of the index weight by the use of expert scoring method,and thus constructs a both qualitative and quantitative index system which may mirror the spreading depth and range of college network public opinions.After the consistency check of the weight value of various indexes,the index value S for network public opinion judgment may be calculated based on the constructed model,and then the warning level may be determined in accordance with the threshold value interval that S belongs to.The further analysis of the variation features of S and mastery of its developmental trends,the revelation of the problem in light of its nature,and thus the prediction of the development of public opinions may help decision makers with correct decisions to guide and control them.Empirical research results show that,based on the collection and organization of network data as well as the rational judgment of experts and decision-makers,the public opinion level may be judged in a timely and accurate way,which may thus provide support for early warning,guidance and intervention so as to effectively control the developmental trend of public opinions.Two typical cases are then worked on,i.e.the bombing in Tianjin Binhai New Area and the school-transferring of 17 postgraduates of Hunan University regardless of rules and regulations.The spreading mechanism of Internet rumors as well as the early warming mechanism and evolution process of network public opinions are studied based on the theoretical research above.The key words of rumors and public health impact are collected to describe the events,and the results of the model are analyzed with characteristic data substituted into it.Findings show that different types of network events have respectively obvious inherent laws and characteristics,and that the analysis and judgment model of network public opinions established in the present study is valid. |
