Statistical hot topics mining and knowledge graph analysis based on LDA model
XIAO Ming1,2， SHANG Huiyu3， XIAO Yi4， LIAO Lili1
(1.Informatization Office， Central China Normal University， Wuhan 430079， China；2.Research Center for Language and Language Education， Central China Normal University， Wuhan 430079， China；3.Digital Intelligence Financial Innovation Lab， Zhougyuan Bank， Zhengzhou 450046， China； 4.School of Information Management， Central China Normal University， Wuhan 430079， China)
Abstract：In order to reveal and compare the development trend and hot topics of CSSCI journals in the field of statistics， a total of 41 495 publications of CSSCI journals in statistics from 1985 to 2020 were collected from CNKI database as the research object. LDA topic model and co-occurrence network model were used to analyze the hot topics and the evolution trend of popular topics and mainstream research methods， and draw the map of relevant knowledge network. Results show that structural equation model and quantile regression method have been widely used in research methods in recent five years， and big data has become a new high-frequency word in recent years. LDA model can accurately mine popular topics and research methods in the field of statistics and provide important support for researchers and decision makers to carry out frontier scientific activities.
2022, Vol. 56 
