Research On The Mechanism Of Top Journal Papers Based On Data Mining And Complex Network

As an emerging research field,the theoretical foundation and research methods of complex networks rely on large-scale interconnected structured data sets.Science of Sci-ence(SciSci),as an extension of complex network theory in the academic world,is a complex network system formed by the self-organization of academic subjects such as scholars,projects,documents,and ideas.SciSci uses the complex network as the the-oretical basis and large-scale academic subjects as the research object by sampling and researching the network system.The topological structure of the network characterizes the development mechanism of the scientific research field itself.It has become another hot research topic in the field of complex science.As far as the previous work in the SciSci field is concerned,its quantitative indicators are practically all based on academic citation and some derived variant indicators,such as yearly citations and the widely used c5,c10 indicator,the total citations of scientific research scholars,the average citations,etc.As one of the most authoritative measure-ment indicators in the scientific research community,the academic citation has shown its extremely high accuracy.However,one of the evident disadvantages of citation data is the strong dependence on the age of documents.The time required for academic works from birth to the saturation of citations is often long and uncertain.Therefore,previous researches are all based on data sets that have reached saturation.But nowadays,this tremendous defect has to be noticed by the further young scholar,because of the fast-changing age.In order to verify the robustness of the experimental results under the premise of avoiding timeliness,this article uses journal influence as a quantitative indicator of the importance of literature.We focus on exploring the mechanism of high-level academic theses and provide some reasonable suggestions for young scholars who are new to sci-entific research.The contents are divided into four parts shown as follows:1.Introduced related concepts and methods in the complex network and SciSci,sum-marized and explained related research directions in the SciSci field,analyzed the current data and theoretical bottlenecks of each research direction,and predicted the future development prospects and application value of this direction.2.Completed the collection of journal influence data based on the data from the major divisions of the Chinese Academy of Sciences,and successfully integrated it with Open Academic Graph(OAG)data,and completed data cleaning progress.3.Explored the influence of the accumulated advantages of academic career on the journal level from different aspects,and realized that the journal level of an article is related to the highest journal-level of all articles published in the author's previous career.4.The random article ranking sequences were used as a comparative experimental group to analyze the appearance position of top journal articles.In the end,it was found that the timing of high-level academic theses appeared randomly in the re-searcher's career articles' sequence.