Research On Topic Distribution Characteristics Of Academic Journals Based On Node Centrality And Balance

Academic journals are one of the important carriers of academic achievements,as well as an indispensable reference resource in the process of knowledge exchange and scientific research innovation.The key words of the papers published in the academic journals are the core content of the papers extracted by the authors,which reflect the subject content of the journal to a certain extent.Analyzing the subjects characteristics of academic journals in one academic area during a period of time is helpful to understand the overall trend of subject development,explore the focus of scientific research,improve the quality and efficiency of scientific research,guide scientific researchers to actively innovate,and promote scientific and technological innovation and progress.Through consulting and combing the relevant literature,we find that the analysis of journal subject distribution is mostly based on the citation relationship and content.Most of the researches based on literature content focus on LDA and co-word analysis to reveal hot topics and topic evolution trend,or use social network analysis to discuss the subject distribution of journals.Based on the existing research,this paper takes the key words of academic journals as the analysis unit,uses the balanced analysis method to measure the difference degree of the central distribution of academic journal subject nodes,and gives an effective method to analyze the characteristics of academic journal subject network and distribution,the specific research is as follows:Firstly,we propose the specific analysis method of the characteristics of academic journal subject network,Gini coefficient is used to measure the equilibrium of journal subject node centrality,we construct the analysis index of central equilibrium,and takes 20 CSSCI journals in the field of library information and archives management in China as an example to test the rationality of the index.The results show that: considering the balance of topic node centrality,CE-has a more comprehensive description of journal subject network structure characteristics;compared with network density,the analysis results of CE-is more consistent with the analysis results of network density,while CE-'s explanation of network characteristics is more detailed.Secondly,we analyze the distribution characteristics of journal subjects on the basis of CE-.Firstly,the clustering method is used to divide the sbject clusters,and then the degree centrality of different subject clusters is calculated;secondly,the Gini coefficient is used to calculate the equilibrium degree centrality of the subject clusters.Finally,according to the number of journal topics(n)and the equilibrium-Degree centrality of journal subject cluster(Cd E-),we construct the two-dimensional coordinate to show and discuss the distribution characteristics of academic journal subjects.Through the empirical analysis,we find that:(1)in the core journals of library science and information science and archives management in China,most of the journals have relatively concentrated subject distribution,and have obvious professional characteristics;(2)the two-dimensional coordinates are more detailed in the classification of the distribution characteristics of the journal subject,and the results are more comprehensive and detailed.Based on the previous research on the subject of journals,this paper proposes a new analysis method combining the node centrality and equilibrium.Based on the network perspective,we analyzes the characteristics of academic journals subject network,and analyzes the distribution characteristics of academic journals from the perspective of content.Taking 20 CSSCI journals in the field of library science and information science and archives management in China as examples,we make an empirical study on the rationality and effectiveness of the new method in order to enrich the theoretical system of periodical subject research.