Data-mining Research On Economic Doctoral Dissertation

The economy of China has experienced a long-term increase since the 1980 s.researchers in academia have also carried out extensive research into economic development which includes the theory of economy,positive economics.Besides,economic research has been extended by scholars,and the extension of research can be reflected in subdivision of research objective,upgradation of research methods,and also has tendency towards modeling and quantification.Positive economics are regarded as a forward movement of studies,but there are still some technical problems such as complicating easier questions,abusing mathematical models,misusing index.Based on the background mentioned above,doctoral dissertations could be more worth referencing and provide more helpful information,compared with journal articles and master's dissertations.This dissertation chooses 12,648 economic doctoral dissertations published from 2000 to 2017 on CNKI website(Zhi Wang).As for research methods in the dissertation,bibliometric,text mining technology,cluster analysis and visualization technology are considered as research methods to analyze what is theme of chosen economic doctoral dissertations,what are methods they choose and what are problems they find.The conclusion of this dissertation includes:(1)economic doctoral dissertations of China gradually become mature,and the number of economic doctoral dissertations submitted to CNSI tends also to be stable.The number of applied economics theses is five times higher than the number of theoretical economics theses,and this might reflect shortages of theoretical economics studies.(2)As for theme of Cluster Analysis,it is worth for researchers to pay more attention to System Dynamics,New Urbanization,the Real and Virtual Economy,Pension and hospitalization insurance,and Medium-sized & Small Enterprises Finance which could represent academic popular areas.(3)the problem of method application is that researchers consider the number of models in dissertations as an important part of high standards of theses,and auditing models and quantitative methods are still main methods that scholars use,but the problem is that chosen models and index or theme of research are sometimes not a perfect match.(4)it is important to research whether methods,index and data we choose in a study are applicable to each other,since sometimes different research methods and chosen index may result in opposite conclusions.For example,there are many problems reflected from GDP and inflation analysis,and this is why abusing and misusing index should get more attention from scholars.