Research On Automatic Evaluation Of Urban Housing Price Based On Characteristic Price Model

With the development of China's market economy,real estate transactions become more frequently. The traditional real estate assessment methods like market comparison approach, income approach, cost approach, can't meet the requirements from financial sector housing lending audit and tax department stocking transaction tax. Real estate value assessment have some problems like large number, tight time and strong consistency. Introducing the automatic assessment can make up for the lack of traditional assessment. This paper summarizes and reviews the current situation of automatic evaluation of domestic and foreign research. Using land prices as an indicator for spatial interpolation to divide the residential sub-market.Quantifying these indicators using direct scoring, indirect scoring, and buffer analysis.Contrasting and analyzing of linear, logarithmic, semi-logarithmic, inverse logarithmic of characteristic price model, and determining effect evaluations.Choosing the best form of them. Getting the automatic evaluation system. Taking Qing Yun as an example for empirical analysis. A total of three residential sub?markets were divided into three residential sub-markets by means of spatial price interpolation, and 12 price characteristics including room age, floor area ratio,distance from commercial center and building structure were selected from three aspects. Four forms of characteristic price model are constructed by using regression analysis. After the comparative analysis, the price model of three residential sub-markets is selected, and the benchmark land price and correction coefficient of 47 residential districts are obtained. Ultimately, establishing the Qingyun County automatic assessment system.The main conclusions include the following aspects:The first, it is more reasonable to divide the residential land market by the spatial interpolation of residential land. Using land price interpolation division residential market compared to the sample price interpolation residential market, which can effectively avoid the mistake like housing age, building structure result on housing prices are not uniform, so that the interpolation division. The former is more reasonable.Second, the construction of various forms of feature price model is helpful to improve the accuracy of automatic evaluation. The use of the same data for four different forms of the characteristic price model will fit out completely different effects. According to the statistical test results, it is the best form of the characteristic price model with the highest fitting degree, which is helpful to improve the accuracy of the automatic evaluation.Third, there are high similarities between the price characteristics of the residential sub - market, but the effect of the specific price factors on the prices of different sub - markets is quite different. Urban housing prices are affected by many factors, the housing prices have a common impact factors, including housing age,building structure,business center and public service facilities. Different housing prices affect the price of the dominant factors,including business center-led, living support and so on.Fourth, Qing Yun County is taken as an example, this paper divides out three residential sub-markets with the average land price of 47 residential districts in the city. Based on the actual situation of Qing Yun County, 12 price characteristics are selected and quantified. The price model of each residential sub-market is established,and the benchmark house price and its correction coefficient of each district are obtained. The automatic evaluation system of residential area in Qing Yun County is formed. The example verifies that the evaluation results meet the requirements.