Online Public Opinion Situation And Threat Assessment Technology Research Based On Intuitionistic Fuzzy Reasoning

With the rapid development of Internet, tremendous freedom is provided for people to express personal opinions. By means of Internet, people can publish their attitudes and opinions for the public affairs concerned. The online public opinion situation and threat assessment analyzes data of the online public opinion topics, which is collected from the internet. The possible online public opinion threat grade is obtained,during the occurring, diffusing and vanishing processes,which offers decision making assistant for related departments to find out and manage public opinion timely and effectively. This dissertation makes deep researches on the function model and key technology of online public opinion situation and threat assessment.Some major contributions are listed as follows:(1) Referring to the idea of situation and threat assessment in military affairs,a function model of the online public opinion situation and threat assessment is designed in this paper.In the model, three phases of the online public opinion situation, including detection, understanding, and forecast, are analyzed based on the data of online public opinion.The online public opinions threat grade is estimated based on the results of the online public opinion situation assessment.(2) The public response can be concluded to be an issue of synthetic judgement oriented to situation understanding. A method used to analyse the public response is proposed based on the intuitionistic fuzzy comprehensive evaluation.Firstly,weights of the two attributes including public quantity and diffuse speed are defined. Secondly,intuitionistic fuzzy evaluation to the public response is defined,and the intuitionistic evaluation matrix is gained.Thirdly comprehensive evaluation of the public response is obtained using intuitionistic fuzzy comprehensive evaluation.Finally,true values are synthesized,and the situation of the public response is analyzed using the principle of maximal true values.The method is proved efficiency and validity through instances.(3) Aiming at correlation among the four attributes: sentiment, attention, sensitivity point and focus, a method used to analysed the relation between topic and the public is presented based on Choquet fuzzy integral.Firstly,intuitionistic fuzzy evaluation to the relation is defined. Secondly, the comparison matrix of shapely values are constructed based on the experience of the experts,and the shapely values of the four attributes are obtained, and the fuzzy measures are obtained by maximizing the Marichal entropy. Thirdly,the comprehensive evaluation of the relation between topic and the public,which different attributes contribute to the concept,is obtained using the Choquet integral.Finally,true values are composed, and the relationship between topics and public are analyzed using the principle of maximal true values. The method is also proved efficiency and validity through instances.(4) Aiming at the auto-classification of online public opinion threat grade,a method based on the intuitionistic fuzzy reasoning is put forwards.Firstly, the conditions and the four threat levels are intuitionistic fuzzyed. Secondly, the nonlinar relationship between the conditions and the threat levels is established according to some rules. Thirdly,the apporach degrees between the result and each threat level are calculated. Finally,the threat level is determined using the maximal apporach degree. Experimental results show that this method is effectively and accuracy.