GIS의 공간통계분석을 활용한 지가분석 = (The) application of spatial statistic analysis with GIS for land value analysis

다국어 초록 (Multilingual Abstract)

The purpose of this paper was to measure the effect of location on evaluating land value with the implementation of GIS coupled with spatial statistical analysis. We assumed that the hedonic price model, which was generally used in this case, could not explain the spatial effect fully. Because the hedonic price model was a location-blind model. We considered that the residual of hedonic price model were relevant to spatial effect and was analyzed to investigate spatial autocorrelation. Also, this research study reviewed the integration processes needed to complement a commercial GIS system to perform spatial statistic analysis.
This objective was been accomplished through the processes such as overlay, interporlation and other spatial analyses. The most of these procedures are used to manage the geostatistic data, which was one of three general classes of spatial data: lattice data, point data, and geostatistic data. The 54 data sampled from 1421 parcel-based land price data were used to predict the unsampled land value according to Kriging which was on the basis of spatial autocorrelation between the locations of vector data. This Kriging process allowed the spatial correlations to be incorporated into the predictions.
We selected the S+ArcView link system as a modular approach. It was the one of recently developed implementation techniques which linked ESDA and GIS by means of a combination of loose or close coupling methods. The distinguishing characteristic of this technique was a close coupling between the S-Plus software as a statistical analysis server and ArcView as the client. It was obtained through a set of specialized functions that take advantage of the S+API and are loaded as an extension to ArcView. This method allowed the user to enter S-Plus commands through a GUI in ArcView and to pass the spatial data to S-Plus for use in the S+Spatialstats module, in which the variogram cloud plot was readily developed.
The result is expected to be applied to the following. We finds that the spatial autocorrelation from a variogram analysis is very useful to estimate or predict the land value associated with spatial context. Namely, the premium and penalities due to spatial neighborhood effect and accessibility can be applied to noting the location, giving a value that includes both structure and neighborhood effects. In addition, it is also expected to determine the similar neighborhood boundary on the basis of being estimated location effects. finally, if ESDA and CSDA is ideally suited to be integrated with GIS, the spatial analysis capacity in GIS is able to be extended.

목차 (Table of Contents)

• I. 서론 = 1
• 1. 연구배경 = 1
• 2. 연구목적 = 2
• 3. 연구방법 = 3
• 4. 논문의 구성 = 4
• II. 이론적 배경과 기술적 연구 동향 = 6
• 1. 지가 이론 = 6
• 1.1 지대와 지가의 개념 = 6
• 1.1.1 농지지대이론 = 6
• 1.1.2 도시지대이론 = 7
• 2. 공간자료와 GIS 공간분석기능 = 10
• 2.1 공간자료의 특징 = 10
• 2.2 GIS의 공간분석 기능 = 11
• 3. GIS의 공간통계분석 = 14
• 3.1 탐색적 자료 분석 = 14
• 3.2 확정적 자료 분석 = 18
• 3.3 GIS 통계분석들의 통합 = 19
• 3.4 GIS 공간통계분석에 대한 기존연구 = 21
• III. 공간통계분석을 통한 지가의 입지 측정 = 25
• 1. 연구지역의 입지 분석 = 25
• 2. 지가의 탐색적 공간자료분석 = 28
• 2.1 지가자료의 선택ㆍ처리 = 28
• 2.1.1 중심적 측정 = 29
• 2.1.2 공간 거리 행렬 = 30
• 2.1.3 특성가격함수를 이용한 입지적 특성값 가정 = 31
• 2.2 지가자료의 분석적 가시화 = 33
• 2.2.1 지가의 공간분포 분석 = 33
• 2.2.2 입지적 특성의 공간 구조 = 38
• 2.3 입지적 특성의 공간적 자기상관관계 분석 = 41
• 2.3.1 입지적 특성의 경험적 변동도 분석 = 41
• 2.3.2 입지적 특성의 이론적 변동도 구축 = 50
• IV. 입지 특성을 고려한 모형화 = 53
• 1. 입지적 특성값 추정을 위한 크리깅 내삽 = 53
• 2. 공간적 특성가격모형의 적용 = 60
• V. 결론 = 64
• 부록 1. 토지특성변수들의 Scatter matrices 분석 = 66
• 참고문헌 = 67
• Abstract = 71