Research On Spatio-temporal Behavior Analysis Of House Buyers Based On Real Estate Data Mining

The ultimate goal of the implementation of the real estate unified registration system is to realize the real estate data information sharing and disclosure.Real estate registration data is a very important part of real estate data,containing important information such as temporal and spatial information.Its unique resource asset attribute has great research and application value,and has important reference value for promoting government and department decision-making,strengthening urban management and maintaining social stability.At present,the research of real estate registration data in China mainly focuses on the integration and construction of real estate registration data,and the research on the application of real estate registration data is still in the initial stage.In the meantime,researches on spatio-temporal behavior of human groups based on industry spatio-temporal data have gradually become a research hotspot in recent years.In view of those phenomenons,this article took house buyers in the real estate registration data as the research object,and conducted research on laws of spatio-temporal behavior of house buyers,which not only expanded the real estate data application field,but also enriched the research content of space-time behavior analysis.Under the guidance of the theory of spatio-temporal behavior,data mining and spatiotemporal data mining theory,and spatial autocorrelation theory,this paper studied the data processing and mining method of real estate registration,and deeply mined laws of spatiotemporal behavior of house buyers in real estate registration data.The research content of this paper mainly includes the following aspects:(1)Clarified the real estate registration data processing method.On the basis of studying the geocoding technology and data cleaning methods,and according to the demand of real estate registration data mining,transformed text containing location information in the real estate registration data into space coordinates,and the real estate registration data was densified and data cleaned against the data quality problems in the real estate registration data.And finally obtained the qualified real estate registration data that meets data mining requirements.(2)Reaearch on the method of data mining of real estate registration.Combined spatiotemporal behavior analysis method,data mining method,spatiotemporal data mining method,GIS method and visualization method,formed a real estate registration data mining method is formed.Through the data mining method of real estate registration,analyzing the behavior preferences of house buyers,the spatio-temporal distribution of houses,and the spatiotemporal behavior of the house buyers in the field.(3)Verified the idea of real estate registration data processing,cleaning and mining by an example.Selected real estate registration data of Dongtai City in Jiangsu Province,Jintan District of Changzhou City and Wujiang District of Suzhou City from 2009 to 2014 to conduct experiments and finding out the rules of spatio-temporal behavior of different types of house buyers.