Research On Spatial Data Mining Method Of China Statistical Yearbook Partition Data

Statistical yearbook data is a kind of important data set,which has the characteristics of authoritative data,timely response,continuous publication,rich content and so on,and has high research value.However,for a long time,researchers still stay in the level of data itself or general statistical calculation,and have not been able to take statistical yearbook data set itself as the research object to mine the hidden knowledge content,It is difficult to find the spatial distribution characteristics and trends of data,the potential association between data,and the spatial distribution relationship between data.Using data mining technology and GIS technology,this paper makes an exploratory research on spatial data mining of statistical yearbook data.Taking the regional statistical data in China Statistical Yearbook as the data set,this paper constructs the data mining model and mining tools,and makes a case study.The research results are as follows:First,the spatial data mining model and tools of statistical yearbook are constructed.In view of the problems existing in the current statistical yearbook data research and the data characteristics of statistical yearbook itself,this paper constructs the data mining model of statistical yearbook from the three directions of data visualization,association rules and spatial relations,and develops the mining tools according to the designed mining model and web technology.Second,through data mining application and case analysis,the validity of the data mining model is verified.Taking China statistical yearbook data as the research data set,this paper conducts spatial data mining from data visualization,association rules and spatial relations to get the mining results knowledge base,and then analyzes the population research as a specific case.Firstly,according to the results of data visualization mining,this paper analyzes the spatial distribution of population in China from three scales of province,city,district and county.It shows that by combining data spatialization with map visualization,more specific spatial distribution pattern,more obvious spatial change trend and more detailed local distribution characteristics of population can be found;Through spatial relationship mining,six factors affecting the spatial distribution of population were found,and the results were evaluated qualitatively and quantitatively by using Lisa map and geographic detector,respectively,to verify the reliability of the results;Through association rules mining,the influencing factors of population quantity are found.Based on the influencing factors,the population prediction model is established and verified by simulation,which shows the effectiveness of the influencing factors.The results show that the data mining model constructed in this paper can effectively mine valuable knowledge from the statistical yearbook data set,and can provide reference for related research and policy-making,so that the statistical yearbook data can be used more and play a greater value.At the same time,this study also provides research ideas for the researchers of statistical yearbook data mining.