Identification Of Research Hotspots In Academic Journals Based On “problem” And “method” Knowledge Element

Research hotspots in academic journals represent the research focus and development direction of subject areas.Identifying research hotspots in academic journal papers helps researchers understand the development history of the field and helps them select research topics accurately and efficiently.In recent years,the research on knowledge organization from the granularity of knowledge elements has been flourishing,fine-grained analysis of text content can extract valuable information from the complex structure of disordered information.However,most of the existing researches are based on macroscopic view or single knowledge element,such as "keywords" and "subject terms",to excavate the hotspots in the field,but few scholars analyze the hotspots of academic journal papers from the perspective of issue and method knowledge elements.This study is based on the extraction of "subject" and "method" knowledge elements of journal papers in the subject area to discover research hotspots in academic journals,and to remedy the problem of coarse granularity in identifying research hotspots.The study is organised around the following aspects:(1)Starting from the developing background of research hotspots and knowledge services of academic journal papers,this research systematically compares the status quo of research hotspots identification and knowledge meta research of academic journal papers through literature research,and establishes the basis of this research.(2)Introducing relevant concepts and research methods,defining the basic concepts of subject and method knowledge elements,illustrating the rule-based knowledge element extraction method,and introducing the construction of knowledge element description rule and steps of domain knowledge element library construction;introducing the concept of LDA topic model and explaining its basic principles.(3)Proposes a method for identifying research hotspots of academic journal papers based on the calculation of issue and method contribution degrees.Using the abstracts of academic journal papers as the data source,the LDA topic model is used to extract the topics of academic journal papers,and the topic names corresponding to the issues and methods are manually labeled,and the Numpy program library is used to establish the "issue-method" knowledge element combination relationship,and the contribution formula of the knowledge element combination is designed,based on which the hot topics of academic journal papers are analyzed.(4)The rule-based knowledge element extraction method is proposed to identify the hot topics of academic journal papers.We establish the description rules for subject and method knowledge element extraction,constructing the subject and method knowledge element database of academic journals,extracting subject and method knowledge elements based on the knowledge element database;using the knowledge element database as a lexicon to extract “subject topics” and “method topics” by using LDA topic model.The “subject” and “method” knowledge elements are extracted based on the knowledge element database.Based on previous studies,this study gives two methods for identifying research hotspots of academic journal papers in terms of knowledge element granularity,and finds through empirical research that the method based on issue and method contribution degree calculation are less workload and corpus is easy to obtain,but the results are somewhat subjective;In contrast,although the preliminary workload is larger,the rule-based knowledge element extraction method is able to subdivide issue and method knowledge elements in scientificity,comprehensiveness and accuracy.The two hotspot identification methods proposed in this study not only enrich and expand the hotspot identification of journal papers,but also help academic journals improve their journal quality and branding;at the same time,they help relevant departments improve their knowledge management and realize value-added knowledge innovation.