Research On The Knowledge Flow In Citation Network From The Perspective Of Citation Content

Academic document is the main form of knowledge output in scientific research activities,reflecting the accumulation of knowledge in the field of scientific research.Academic documents do not exist in isolation.One academic document is usually based on multiple academic documents,and the inheritance and change of knowledge are realized by citation behavior.Citation analysis is a scientific metrological method based on the citation relationship of academic documents,which is used to reveal its quantitative characteristics and internal laws.The early citation analysis focused on the frequency of citation of academic documents,measured the quality and value of academic documents by the number of citations,and proposed many methods of scientific knowledge evaluation based on citation network analysis.In fact,citation network analysis reveals the innovative relationship between academic documents from the perspective of knowledge flow,which has important functions such as scientific knowledge evaluation,scientific development mode disclosure and scientific frontier exploration.However,the traditional citation network ignores the semantic information in citation content,does not explore the knowledge association between academic documents,and cannot reveal the location,motivation and strength of citation.In citation network analysis,all citation relationships are treated equally,but the value,contribution and importance of the cited documents to the citing documents are not exactly the same.In addition,the complexity of citation behavior poses a great challenge to the analysis of citation network only considering the frequency of citation.These defects also make the research on knowledge flow in citation network a bottleneck.With the development of citation analysis,the importance of citation content analysis has been paid more and more attention.The analysis of citation content is based on the internal characteristics of academic documents,extracting the subject content of the cited documents from the objective texts of the cited documents,and then revealing the innovative relationship between the cited documents and the cited documents as well as the reasons for the cited documents.With the development of open access movement,it is very convenient to obtain full-text data of academic literature.At the same time,the text structure recognition technology represented by citation location acquisition and text content mining technology represented by citation subject and citation intensity calculation also provide a broader space and possibility for academic literature analysis and measurement research based on full-text data.In view of this,this paper first redefines the construction of citation network from the perspective of citation content,making it a heterogeneous,weighted and multidimensional network with different number of knowledge units and different weights and types of knowledge links(edges);Secondly,this paper select descriptive statistical indicators and network analysis indicators to describe and analyze the knowledge flow ability of knowledge nodes;Then,this paper divide the whole citation network and explore the knowledge flow characteristics in citation network from different knowledge attribute dimensions,as well as the knowledge flow differences between different range of citation networks.Finally,the simulation model is used to explore the main reasons that affect the knowledge flow of citation network,and the main factors that affect the knowledge flow in citation network are obtained by comparing and analyzing the results between simulation experiment and real experiment.This paper is divided into seven chapters:The first chapter mainly discusses the research background of knowledge flow in citation network from the perspective of citation content,the research status at home and abroad,the research ideas and technical routes of this paper,the innovation of this paper and the organizational structure of this paper.The second chapter describes the theoretical basis of this paper,including citation analysis theory,network analysis theory,knowledge flow theory.Through clarifying the relationship among knowledge flow,knowledge network,citation network and citation content,this paper determines the research ideas of constructing citation network based on the analysis results of citation content,and then carrying out the research on citation network knowledge flow.The third chapter,the fourth chapter,the fifth chapter and the sixth chapter are the core chapters of this paper,which are carried out according to the specific research process of "knowledge flow model construction-knowledge flow network analysis-knowledge flow characteristics analysisknowledge flow effect analysis".The third chapter is the construction of knowledge flow model in citation network from the perspective of citation content.This chapter integrates the results of citation content analysis among academic documents into the traditional citation network construction process,redefines the knowledge flow model and its constituent elements in citation network,realizes the disclosure of knowledge association,transformation and innovation in citation network from the content level of the citation,and lays the foundation for the analysis of rules and patterns of knowledge flow in citation network in the next stage.The fourth chapter is the analysis of knowledge flow network in citation network from the perspective of citation content.This chapter selects descriptive statistics and network analysis indicators to carry out multidimensional detection on the knowledge flow characteristics of citation networks integrated with citation content,and conducts comparative research on the results of citation network analysis in different subject areas,so as to extract more universal rules and patterns of knowledge flow in citation network,mainly including in-depth description and analysis of three parts,respectively knowledge flow ability and role of knowledge node,knowledge flow type and structure of knowledge community,knowledge flow distribution and structure characteristics of the whole network.The fifth chapter is the analysis of knowledge flow characteristics in citation network from the perspective of citation content.This chapter extracts the individual citation network reflecting different knowledge attributes from the whole citation network,and closely combines the characteristic data of knowledge nodes with the relational data of knowledge nodes,so as to reveal the knowledge flow characteristics in the citation network from different dimensions and in a deeper level.The main method is to obtain and disassemble the information such as title,author,organization,publication name,year and keyword in the citation data,and to integrate and refine citation network from four dimensions,including literature,domain,region and theme,and to introduce time series to analyze the knowledge flow characteristics of citation network.The sixth chapter is the analysis of the knowledge flow effect in citation network from the perspective of citation content.This chapter constructs the evolutionary dynamic model of citation network knowledge flow from two aspects,including static characteristic attribute and dynamic structure attribute,and uses Netlogo,a simulation software,to generate different types,characteristics and attributes of citation network that integrate the results of citation content analysis.On this basis,this chapter constructs the measurement method of knowledge flow effect function and its effect coefficient from three aspects,including knowledge absorption effect,knowledge superposition effect and knowledge cost effect,and uses real data to design real experiment to verify the effectiveness of simulation experiment results.The seventh chapter is summary and prospect.On the one hand,the research work and content of this paper are summarized.On the other hand,the insufficient parts and future directions of this paper are discussed.The results in this paper show that exploring the knowledge flow phenomenon in citation network from the perspective of citation content can reflect the characteristics of knowledge flow among academic documents in citation network with more objective analysis results,make up for the main defects caused by neglecting the deep-seated citation information in the study of knowledge flow in citation network in the past,and play an important role for revealing the knowledge structure,knowledge association and knowledge evolution,and optimizing the evaluation of quality and influence of academic documents.