Research On Emotional Semantic Space Construction And Emotional Semantic Intensity Measurement Of Internet Public Opinion

Under the background of mobile Internet and big data,network public opinion has played an unprecedented important role in expressing public sentiment.Under the environment of big data of network resources,with the popularization of various new network services and the strengthening of citizens' awareness of participation,network public opinion has diffused complexity.And diversification has reached unprecedented levels.Internet users have great differences in social experience,information literacy and education,and at the same time,influenced by identity secrecy,public opinion information representing extreme or irrational views is more likely to spread.Faced with the vast amount of information,how to identify the emotional tendencies contained in it,identify and integrate the semantics of the emotional information of network public opinion,so as to penetrate the real attitudes and opinions of users to social hot events,and then effectively monitor,warn and guide public opinion is the reality of monitoring public opinion crisis.Demand is also a key problem that must be resolved to enhance public opinion crisis assessment and response capability.On the basis of systematically combing the research results of network public opinion sentiment analysis both at home and abroad,this paper comprehensively analyzes the connotation,characteristics,measurement and relevance of semantic space dimension of network public opinion by using literature analysis,social network analysis,empirical research and Simulation Research methods.Based on the comprehensive analysis of the concept and hierarchy of emotional semantic elements,which are the basic units in sensory semantic space,an emotional semantic space of network public opinion is constructed.An Empirical Study on the model of emotional semantic space proposed in this paper is carried out to further express and quantify the characteristics of emotional semantic space of network public opinion.And evolution rules,which provide a basic theoretical and technical reference for the research of affective computing in online public sentiment.The first chapter is the introduction,which mainly expounds the era and historical background of the development of network public opinion information,expounds the purpose and significance of this study,summarizes the current situation of the research on network public opinion information emotion at home and abroad,and finds out the key problems as the basic starting point of this study,and analyzes the methods and technical approaches used in this study.Finally,the main research contents and research innovations of this paper are summarized.In the second chapter,the related theories are summarized,including information theory and full information theory,basic emotional theory related to psychology and information psychology,discrete emotional theory,emotional dimension theory,emotional prototype theory,etc.At the same time,the measurement methods of network public opinion information analysis are analyzed and compared.Finally,the development of the theory and technology of emotional computing and emotion recognition and the application of this study are described.The third chapter is the analysis of the spatial semantic dimension of Internet public sentiment.Based on the theory of emotional hierarchy,this chapter analyzes the semantic characteristics of network public opinion,puts forward the concept of information emotional dimension of network public opinion,and defines the hierarchical framework of each dimension of network public opinion emotion.This paper analyzes the emotional elements in the semantic space of network public opinion,explains the concept and hierarchy of the emotional semantic elements of network public opinion,analyzes the relationship between the emotional semantic elements of network public opinion and the dimension space of emotion,and then puts forward the methods and key technologies for identifying the semantic features of network public opinion,and analyzes the emotions of participants in network public opinion.The process of perceptual semantic cognition and transformation provides a useful reference for the hierarchical monitoring and crisis warning of network public opinion information.The fourth chapter is the space construction of the emotional dimension of network public opinion information.Based on the previous research foundation of the emotional dimension of network public opinion information,the concept,constitution and characteristics of the emotional dimension space of network public opinion information are further put forward.The construction principle of the emotional dimension of network public opinion information is put forward,and the emotional dimension of network public opinion information is explained in detail.Based on the above analysis,the PAD psychological model is applied to construct the emotional dimension space of network public opinion information,and then the mechanism of multi-dimensional emotional synthesis in the emotional dimension space is elaborated,and the emotional space movement mode of network public opinion information is summarized.The fifth chapter is the construction of emotion ontology and mapping of semantic space.Ontology,as the carrier of systematic description of domain knowledge,can simulate human cognitive situations and effectively solve the problems in emotional semantics.Ontology can standardize concepts and terms in the emotional domain and depict the inner links between concepts.The semantic characteristics of ontology are used to connect emotional features with high-level semantics.Meanwhile,the scope and depth of emotional semantic representation are further expanded by the reasoning ability of ontology,so that the emotional computation based on ontology can more accord with the law of public opinion development.The sixth chapter focuses on the analysis of Internet public opinion hotspots based on emotional semantic space.Based on the theory of affective semantic space,this chapter applies LDA clustering method and time series analysis method to innovatively propose an efficient and feasible public opinion analysis model--micro-blog public opinion analysis model based on affective semantic space,through topic clustering of LDA in text mining and intensity measurement based on affective semantic space.The topic sequence is calculated,and then the time series equation of topic emotion intensity is constructed.Through regression analysis,the trend of public emotion in the process of event development is analyzed,and the hot spot of public opinion is tracked.Finally,the empirical analysis not only proves the feasibility of this method,but also opens up a new way for hot spot analysis of public opinion.The seventh chapter is the wom evaluation model based on emotional semantic space.In this chapter,a method of monitoring and evaluating online word-of-mouth based on network emotional semantic space is proposed.Using ACSI model,the online word-of-mouth model is divided into four dimensions: corporate image,perceived expectation,perceived value and perceived quality.The emotional data set of HowNet,the extended version of synonym forest and the emotional lexicon ontology of Dalian Polytechnic are constructed.An emotional dictionary was built to map the emotional feature words in the comments into the PAD emotional space,thus realizing the refinement of consumers' emotions.A set of theoretically feasible and practically simple online word-of-mouth monitoring and evaluation method was constructed.The eighth chapter summarizes and prospects,summarizes the research work,innovative achievements and shortcomings,and gives the idea of further research.The emotional semantic content of network public opinion is a relatively complete emotional space system with implicit,multi-dimensional and relevance.It explores the internal law of emotional data in network public opinion information and the scope of multi-dimensional information emotional data to provide useful reference for the efficient guidance of network public opinion and to enhance public opinion?the level and ability of crisis prediction,management and guidance.