Research On Construction Of User Topic Graph In Social Networks And Guiding Strategy

In the report of 19 th CPC National Congress,General Secretary Xi Jinping issued a great call to the whole Party and the whole nation for "consolidating confidence towards our culture and promoting the prosperity of socialist culture"."Overall national security concept" is also regarded as the basic strategy of building socialism with Chinese characteristics under the background of new era.Effective supervision of public opinion in social networks is an inexorable requirement to cope with the new changes and new development of national security environment in the new situation.However,the virtuality of the network brings huge difficulty to the public opinion supervision.They're effective breakthrough points for public opinion supervision to build a social network user topic graph system model by effectively using the knowledge graph,explore the potential user communities,confirm the identity characteristics of users,and work out the emotional tendency of users.Combing with methods like literature analysis,empirical research,knowledge graph,machine learning,etc.this paper builds the social network public opinion user topic graph,and puts forward guidance strategies as well.Specifically speaking,this paper primarily includes six parts.Firstly,the third chapter analyzes the social network user topic graph system model,and it is the theoretical core framework of the whole paper,which points out the four key elements of the social network user topic graph system model,such as information environment,information people,information and information technology,and forms a multi-dimensional topic graph through user community graph,user identity graph and user emotion graph.Secondly,from Chapter 4to Chapter 6,taking the "air crash of Egyptair" as a case of typical topic,this paper analyzes the discovery of social network public opinion user community,user identity classification and emotional evolution rules through empirical analysis method based on the information environment,information people and information respectively,which provides theoretical support for Chapter 7 and Chapter 8.Next,the Chapter 7 analyzes the ecological evaluation of social network public opinion and establishes the ecological evaluation index of social network public opinion based on the analysis of Chapter 4,5and 6,which also provides theoretical support for Chapter 8.Finally,Chapter 8 puts forward the guidance strategy of social network public opinion,which is the foothold of this paper at the practical level.The specific contents are shown below.Chapter 3 The construction of social network public opinion user topic graph system.Firstly,this chapter puts forward the information ecology elements of social network public opinion user topic graph,analyzes the environmental elements,subject elements,object elements,and technical elements of the social network public opinion user topic graph.Secondly,combining with the related knowledge of graph technology related theories,it puts forward the entity identification,relationship extraction,attribute extraction and model construction of social network public opinion user topic graph.Next,this chapter determines the analysis model of social network public opinion user topic graph,points out that the social network public opinion user topic graph is composed of three dimensions,such as user community graph,user identity graph and user emotion graph,and goes deep into the community discovery from the information environment perspective,the user classification from the information people perspective,and emotional evolution from the information perspective.And on this basis,conduct an ecological evaluation of the public opinion of the entire social network.Finally,the social network public opinion user topic graph system model is proposed.Chapter 4 The construction of social network public opinion user community graph and the discovery of relationship.This chapter is based on the theory of information environment,combined with the LDA topic model of JS divergence,to build social network public opinion user community graph model,and to discover community relationship.Using the empirical research method,the topic modeling of the public opinion text under the topic of "Egypt Aviation Difficulty" is used,the perplexity index is used to determine the optimal number of public opinion users,the JS divergence is used to measure the similarity,and the calculation result is used as the edge weight.Use VOSviewer software to build the social network public opinion user community graph,further divide multiple network communities,analyze and discuss the theme preferences and user characteristics of the network community,and accurately locate the opinion leaders in the network community.This chapter mainly studies the information environment in social network public opinion,which echoes Chapter 5 and 6,and provides theoretical support for Chapter 7-the ecological evaluation index of social network public opinion(B1,B2),and Chapter 8-the guidance strategy of social network public opinion community from the perspective of information ecology.Chapter 5 The construction of social network public opinion user identity graph and the identification of users.This chapter is based on the information person theory,combined with the LDA topic model and the naive Bayes classifier model to build social network public opinion user identity graph,and identify the user identity.Firstly,LDA topic model is adopted to divide topics,explore the deep semantic features of the texts that public opinion users forwarded and commented,and conclude public opinion user identity features and communication features.Then,naive Bayes classifier is adopted to divide public opinion user types,and Neo4 j is adopted to draw user identity graph.The result shows that there are key persons existing in the public opinion communication from network users of different types,and systematically analyze the social network public opinion users' thematic concerns and evolution process.This chapter mainly studies the information people in social network public opinion,which echoes with Chapter 4 and 6,and provides theoretical support for Chapter 7-the ecological evaluation index(B3,B4,B5)of social network public opinion and Chapter8-the guidance strategy of social network public opinion users.Chapter 6 The construction of social network public opinion user emotion graph and the evolution of emotions.This chapter is based on information theory,constructing a social network public opinion user emotion graph based on word vector-based multi-scale convolutional neural network,and effectively dividing public sentiment user emotional tendencies.Firstly,this chapter builds network user emotion classification model through convolution neural network model,verifies and analyzes the accuracy of the emotion classification established for network users,and verify the superiority of the emotion classification model built for network users.Secondly,Gephi is adopted to analyze the visualization of user emotion graph based on the classification result of the convolution neural network model.This chapter mainly studies the information in social network public opinion,which echoes Chapter 4 and 5 and provides theoretical support for Chapter 7-the ecological evaluation index(B6)of social network public opinion and Chapter 8-the guidance strategy of social network public opinion emotion.Chapter 7 The ecological evaluation of social network public opinion.This chapter establishes the ecological evaluation index of social network public opinion from four dimensions of information environment,information people,information and information technology based on the perspective of information ecology.The comprehensive fuzzy evaluation method is adopted to conduct ecological evaluation on the outbreak stage of the topic of "Egyptian aviation disaster",and the evaluation results are analyzed,which effectively solves the difficult problem of ecological evaluation indicators that are difficult to quantify,and provides ecological evaluation The operability.This chapter echoes the Chapter 4,5 and 6,and provides theoretical support for Chapter 8-the guidance strategy of social network public opinion emotion.Chapter 8 The guidance strategy of social network public opinion.This chapter puts forward the guidance strategy of social network public opinion based on the research conclusions of Chapter 4,5 and 6,as well as the ecological evaluation index of social network public opinion in Chapter 7.Firstly,it raises out the challenges and existing problems from the perspective of ecology in Internet and social network public opinion.Then it puts forward different guidance strategies from three perspectives of information environment,information people and information.Specifically speaking,this chapter proposes to enhance the monitoring of derived topics,improve the promotion of social network public opinion topic and attach importance to the guidance strategy of community service from the perspective of information environment.It also puts forward the guidance strategy to improve the user type management for social network public opinion users,play the role of main streaming media and improve the communication mechanism of opinion leaders from the perspective of information people.Finally,it proposes the guidance strategies such as the guidance on social network public opinion emotion,the fine management of social network content,attaching importance to the guidance of public opinion emotion,and improving the early warning mechanism for public opinion emotion from the perspective of information.This paper puts forward the construction method of social network public opinion user topic graph system at the theoretical level,which is the specific and instantiated application of information ecology theory in the social network field.It builds the social network public opinion user topic graph system model from three different dimensions of user community,user identity and user emotion,so as to enable the controllers to deeply explore the potential user communities,effectively recognize the users' identities and accurately analyze the emotional tendency of users.Meanwhile,the proposed ecological evaluation index provides a unified evaluation standard for social network platforms,as well as a strong theoretical support for the scientific management of social network public opinion.In practice,it can lead the controllers to formulate corresponding control strategies from the three aspects of communities,users and emotions,which can guide the direction of system development for social network platform,guarantee free speech right for users on social network platforms,and promote the healthy development of information ecology in social network as well.