Interdisciplinary Research Based On Journal Level And Paper Level Classifications Of Science

Current interdisciplinary research from a quantitative perspective is mainly based on journal-level classification of science because the current category classification of most English academic database resources is limited to the journal level.It is doubtless that imposing the classification code of a journal on the papers published on the journal will cause errors.Paper-level classification of science can accurately represent the discipline attributes of the papers instead.Fortunately,current discipline classification of mainstream Chinese academic database resources covers the journal and paper levels,providing an opportunity for interdisciplinary research based on paper level classifications of science.Therefore,the Chinese Library Classification(CLC)and citation data of papers published in 2008-2018 from 1769 journals indexed by CSSCI and CSCD is collected.An interdisciplinary research based on journal-level and paper-level classifications of science is conducted.We then compared the results based on the two research perspectives mentioned above,the results we find in this study and existing research.Some differences and commonalities are reflected.Also,not only is the interdisciplinary research based on the classification of paper level more accurate,but also reveal deeper interdisciplinary relationships.Firstly,it is proved that knowledge exchanges among tier-1 categories are very frequent.In the tier-2 categories interdisciplinary research,we can see there are a large number of papers in Metallurgical Industry,Physics and General Industrial Technology related to other disciplines.Science and Scientific Research,Education,Information and Knowledge Dissemination,Automation Technology and Computer Technology are highly interdisciplinary.Also,most tier-2 categories acquire most knowledge from specified disciplines.In the research at tier-3 and tire-4 categories,it is found that computer and automation-related categories such as Computer Technology,Computer Application,and Automation Technology cooperate with the most foreign categories.The betweenness centralities and varieties of them are very high.At the same time,they are also important hubs for knowledge communication between natural sciences and social sciences.In network analysis based on the tier-4 classification,the important position of Artificial Intelligence is also highlighted.A comparative study based on the results of different research methods shows that interdisciplinary studies based on journal-level and article-level classification of science share a lot of similarities,which is reflected by the similarities of the top-ranked categories and the similarities of the core discipline knowledge communication network.Correlation test also shows that the results measured by the two methods have a certain consistency to some extent.However,there are also differences among the results based on different research methods.For example,the relative rankings of disciplines at different tiers are different,and the number of clusters and cluster topics are different too.Interdisciplinary research based on the paper-level classifications of science is advantageous but also limited comparing to it that based on the paper-level classifications.