지원사업
학술연구/단체지원/교육 등 연구자 활동을 지속하도록 DBpia가 지원하고 있어요.
커뮤니티
연구자들이 자신의 연구와 전문성을 널리 알리고, 새로운 협력의 기회를 만들 수 있는 네트워킹 공간이에요.
논문 기본 정보
- 자료유형
- 학위논문
- 저자정보
- 지도교수
- 백태경
- 발행연도
- 2017
- 저작권
- 동의대학교 논문은 저작권에 의해 보호받습니다.
이용수16
이 논문의 연구 히스토리 (3)
- 이 논문의 후속연구가 궁금하신가요?
- 연관 학술논문 또는 학술발표를 통해 보다 발전된 연구결과를 확인하실 수 있습니다.
- 이 논문의 연구 히스토리 확인하기
초록· 키워드
오류제보하기
도시의 인구집중 현상은 다양한 도시문제를 초래하게 되었다. 특히, 도시범죄는 시민들의 불안감을 가중시켜 삶의 질을 악화시키는 원인이 되고 있다. 환경범죄학자들에 의하면 도시범죄는 공간적 특성과 패턴이 존재하며 발생지역의 환경적 요인과 밀접한 관계를 가진다고 주장하면서 범죄예방환경설계 (CPTED: Crime Prevention Through Environmental Design)의 중요성을 제시하고 있다. 본 연구는 범죄와 환경과의 관계가 밀접한 영향을 미친다는 전제하에 범죄억제를 위한 정책수립 및 설계를 위한 대안마련의 사전단계로 범죄위험지역 분석과 요인, 공간적 특성이 반영된 예측모델을 제시하는 것을 주요 목적으로 수행하였다.
기존 범죄관련 연구는 대부분 포괄적인 지역 단위를 대상으로 연구분석을 수행하고 있어 국지적 영역의 범죄발생특성 및 요인에 대한 결과를 제대로 제시하지 못하고 있다는 단점이 있다. 따라서 소규모 지역단위의 분석을 위한 공간적 이질성 측면에서 지리가중회귀분석 (GWR:Geographically Weighted Regression)의 필요성을 제시하고 있으나 실제 적용된 사례가 없다.
따라서 본 연구는 해운대에서 발생한 범죄자료 (2009~2013)를 토대로 다음과 같은 차별성을 두고 수행하였다.
첫째, 실제 범죄발생의 위치자료를 토대로 실제 범죄발생지역의 핫스팟을 도출하였다.
둘째, 기존 연구의 분석단위에서 발생하는 분석의 수정가능성 문제를 줄이고, 신뢰 높은 설명력을 가지기 위해 도시최소 경계단위인 센서스를 기반으로 수행하였다.
셋째, 기존의 범죄발생 영향요인과 관련한 선행연구는 분석의 단위가 포괄적이므로 기존의 예측모델 방식의 한계를 극복하기 위해 공간적 상호작용을 반영한 공간회귀모델링을 적용하여 범죄의 종속성과 이질성 파악에 주력하였다.
주요 분석내용으로 범죄발생지역의 현황 및 패턴, 분포 특성을 점, 면 사상 자료를 토대로 전역적, 국지적으로 구분하여 분석하였다. 또한 범죄발생에 영향에 미치는 요인 및 적합한 예측모형을 제시하기 위해 복수의 공간회귀분석 (다중선형회귀분석 및 지리가중회귀분석) 결과 값을 비교·분석하여 개선된 모형을 도출하였다.
연구의 분석결과, 해운대 전체 지역에 범죄가 균등하게 분포하였을 경우와 대비하여 약 1/4 수준으로 범죄발생지역은 군집성을 나타내고 있었으며 주로 해운대 해수욕장 배후 상업지역에서 특정범죄 (강간, 절도, 폭력)는 강한 핫스팟을 나타내었다.
범죄발생의 영향요인 분석결과 외국인 수, 아동인구 수, 노령자 수, 고학력자 수, 독거여성 수, 1인 가구 수, 풍속업소 수, 치안센터 수, CCTV 설치 수의 9개 변수에서 유의미한 결과를 나타내었다.
전역적 단위와 국지적 단위의 특성을 반영한 기준모형(다중선형회귀모형)과 대안모형(지리가중회귀모형)을 통해 범죄사고의 설명력을 비교한 결과 지리가중회귀모형의 설명력은 0.714로 다중선형회귀모형 0.626과 비교하여 약 9% 가량 적중률이 높은 것으로 나타났고 표준화된 평균제곱근오차 (SRMSE: Standardized Root Mean Square Error)를 통해 예측값에 대한 모형검증 결과 지리가중회귀분석을 통한 모형이 우수한 것으로 나타났다.
결과적으로 다중선형회귀모형을 이용할 경우 일부 지역에서 범죄발생 위험등급이 과대 또는 과소 추정되었음을 알 수 있는데 이 같은 결과는 독립변수와 종속변수의 공간적 상관관계를 고려하는 지리가중회귀모형을 이용함으로써 보다 효과적으로 적용할 수 있음을 알 수 있었다. 예를 들어, 폐공가 수의 경우, 다중회귀모형에서 종속변수와 부적 (-) 관계를 가져 폐공가 수가 증가할수록 범죄 발생 확률이 감소하는 것으로 나타났다. 그러나, 지리가중회귀 분석 결과를 살펴보면, 연구지역 내 회귀계수의 약 52%는 음 (-)의 값, 약 47%는 양 (+)의 값을 가졌다. 이러한 결과는 다중선형회귀 모형의 결과에서와 같이 연구지역 내 폐공가 수가 종속변수와 단순히 부적의 관계를 가지는 것이 아니라, 지역에 따라 종속변수에 서로 다른 영향을 미치는 것을 알 수 있다.
기존 범죄관련 연구는 대부분 포괄적인 지역 단위를 대상으로 연구분석을 수행하고 있어 국지적 영역의 범죄발생특성 및 요인에 대한 결과를 제대로 제시하지 못하고 있다는 단점이 있다. 따라서 소규모 지역단위의 분석을 위한 공간적 이질성 측면에서 지리가중회귀분석 (GWR:Geographically Weighted Regression)의 필요성을 제시하고 있으나 실제 적용된 사례가 없다.
따라서 본 연구는 해운대에서 발생한 범죄자료 (2009~2013)를 토대로 다음과 같은 차별성을 두고 수행하였다.
첫째, 실제 범죄발생의 위치자료를 토대로 실제 범죄발생지역의 핫스팟을 도출하였다.
둘째, 기존 연구의 분석단위에서 발생하는 분석의 수정가능성 문제를 줄이고, 신뢰 높은 설명력을 가지기 위해 도시최소 경계단위인 센서스를 기반으로 수행하였다.
셋째, 기존의 범죄발생 영향요인과 관련한 선행연구는 분석의 단위가 포괄적이므로 기존의 예측모델 방식의 한계를 극복하기 위해 공간적 상호작용을 반영한 공간회귀모델링을 적용하여 범죄의 종속성과 이질성 파악에 주력하였다.
주요 분석내용으로 범죄발생지역의 현황 및 패턴, 분포 특성을 점, 면 사상 자료를 토대로 전역적, 국지적으로 구분하여 분석하였다. 또한 범죄발생에 영향에 미치는 요인 및 적합한 예측모형을 제시하기 위해 복수의 공간회귀분석 (다중선형회귀분석 및 지리가중회귀분석) 결과 값을 비교·분석하여 개선된 모형을 도출하였다.
연구의 분석결과, 해운대 전체 지역에 범죄가 균등하게 분포하였을 경우와 대비하여 약 1/4 수준으로 범죄발생지역은 군집성을 나타내고 있었으며 주로 해운대 해수욕장 배후 상업지역에서 특정범죄 (강간, 절도, 폭력)는 강한 핫스팟을 나타내었다.
범죄발생의 영향요인 분석결과 외국인 수, 아동인구 수, 노령자 수, 고학력자 수, 독거여성 수, 1인 가구 수, 풍속업소 수, 치안센터 수, CCTV 설치 수의 9개 변수에서 유의미한 결과를 나타내었다.
전역적 단위와 국지적 단위의 특성을 반영한 기준모형(다중선형회귀모형)과 대안모형(지리가중회귀모형)을 통해 범죄사고의 설명력을 비교한 결과 지리가중회귀모형의 설명력은 0.714로 다중선형회귀모형 0.626과 비교하여 약 9% 가량 적중률이 높은 것으로 나타났고 표준화된 평균제곱근오차 (SRMSE: Standardized Root Mean Square Error)를 통해 예측값에 대한 모형검증 결과 지리가중회귀분석을 통한 모형이 우수한 것으로 나타났다.
결과적으로 다중선형회귀모형을 이용할 경우 일부 지역에서 범죄발생 위험등급이 과대 또는 과소 추정되었음을 알 수 있는데 이 같은 결과는 독립변수와 종속변수의 공간적 상관관계를 고려하는 지리가중회귀모형을 이용함으로써 보다 효과적으로 적용할 수 있음을 알 수 있었다. 예를 들어, 폐공가 수의 경우, 다중회귀모형에서 종속변수와 부적 (-) 관계를 가져 폐공가 수가 증가할수록 범죄 발생 확률이 감소하는 것으로 나타났다. 그러나, 지리가중회귀 분석 결과를 살펴보면, 연구지역 내 회귀계수의 약 52%는 음 (-)의 값, 약 47%는 양 (+)의 값을 가졌다. 이러한 결과는 다중선형회귀 모형의 결과에서와 같이 연구지역 내 폐공가 수가 종속변수와 단순히 부적의 관계를 가지는 것이 아니라, 지역에 따라 종속변수에 서로 다른 영향을 미치는 것을 알 수 있다.
목차
제1장 서 론 ·······················································································11. 연구의 배경 및 목적 ·································································31) 연구의 배경 ·····························································································32) 연구의 목적 ·····························································································52. 연구의 범위 ·················································································63. 연구의 구성 ·················································································7제2장 이론 및 선행 연구고찰 ·······················································91. GIS와 범죄 분석 ······································································ 111) GIS의 정의 및 구성요소 ····································································112) GIS를 활용한 범죄지리분석 ······························································132. 공간통계분석 이론 ···································································173. 범죄공간통계분석과 관련한 선행연구 ································· 181) 범죄발생 공간분포 연구 ·····································································192) 범죄발생 요인에 관한 연구 ·······························································193) 범죄발생 예측 모형 구축에 관한 연구 ···········································204. 소결 ·····························································································231) 선행연구 동향 ·······················································································232) 연구의 차별성 ·······················································································23제3장 분석방법 및 변수의 선정 ·················································271. 분석방법 및 모형 ·····································································291) 분석 방법 ·······························································································292) 분석 모형 ·······························································································31(1) 다중선형회귀 모형 ··········································································31(2) 지리가중회귀 모형 ··········································································322. 자료의 구성 및 구축 ·······························································361) 자료의 분석단위 설정 ·········································································362) 자료의 구성 및 수집 ···········································································383) 자료의 구축 ···························································································39(1) 집계구 데이터의 구축방법 ····························································40(2) 비집계구 데이터의 구축방법 ························································413. 변수의 선정 및 설명 ·······························································421) 변수의 선정 ···························································································422) 변수의 설명 ···························································································44(1) 인구학적 변수 ··················································································44(2) 사회경제학적 변수 ··········································································45(3) 물리환경적 변수 ··············································································47(4) 방어기제 변수 ··················································································484. 소결 ·····························································································50제4장 범죄발생 현황 및 공간적 분포 특성 ····························· 531. 범죄발생의 현황 분석 ·····························································551) 범죄발생 현황 ·······················································································55(1) 범죄유형별 발생현황 ······································································55(2) 지역별 범죄발생 현황 ····································································57(3) 시간대별 범죄발생 현황 ································································60(4) 장소별 범죄발생 현황 ····································································622) 사회지표별 현황 ···················································································64(1) 인구분포 특성 ··················································································64(2) 사회경제적 특성 ··············································································66(3) 물리환경적 특성 ··············································································69(4) 방어기제 특성 ··················································································692. 범죄발생의 점(點) 사상 공간적 분포패턴 ·························· 711) 점(點) 사상 자료의 분석 개요 ··························································712) 점(點) 사상 자료의 전역적 분석 ······················································73(1) 모드 (Mode) 분석 ···········································································73(2) 표준편차타원체 (Standard Deviation Ellipsoid) ······················74(3) 커널 밀도 추정 (Kernel Density Estimation) ···························75(4) 최근접 거리분석 (Nearest Neighbor Distance Analysis) ·······773) 점(點) 사상 자료의 국지적 분석 ······················································80(1) 최근접 계층군집기법 (Nearest Neighbor Hierarchical) ··········80(2) STAC (Spatial and Temporal Analysis of Crime)분석 ········· 82(3) K-평균 분할군집화 (K?Means Partitioning Clustering)기법 833. 범죄발생의 면(面) 사상 공간적 분포패턴 ·························· 841) 면(面) 사상 자료의 분석 개요 ··························································842) 면(面) 사상 자료의 전역적 분석 ······················································85(1) Global Moran''s I ·············································································85(2) Getis-Ord Local Gi* ········································································863) 면(面) 사상 자료의 국지적 분석 ······················································91(1) Local moran’s I (Cluster and Outlier Analysis) ·····················914. 범죄발생지역의 유형화와 특성 분석 ··································· 961) 용도지역에 따른 범죄발생 특성 분석 ·············································96(1) 공간별 유형화 방법 ········································································96(2) 용도지역별 유형 분류 ····································································97(3) 용도지역 유형별 특성 종합 ··························································992) 계층적 군집분석에 따른 범죄발생 특성 분석 ·····························102(1) 계층적 군집분석에 따른 공간별 유형화 방법 ························ 102(2) 계층적 군집분석에 따른 유형화 지표의 선정 ························ 103(3) 계층적 군집분석에 따른 유형분류 ············································104(4) 계층적 군집분석에 따른 유형별 특성 분석 ····························106(5) 계층적 군집분석에 따른 유형별 특성 종합 ····························1105. 소결 ···························································································112제5장 범죄발생예측 모형의 구축 ·············································1171. 회귀분석을 위한 사전 분석 ·················································1191) 다중공선성 분석 ·················································································1192) 사용된 변수의 기술통계 ···································································1242. 다중선형회귀 (OLS) 모형 구축 ···········································1251) 다중선형회귀분석 결과 ·····································································1252) 다중회귀모형의 적합성 검정 ···························································1283. 지리가중회귀 (GWR) 모형 구축 ········································ 1324. 모형별 분석결과 비교 ···························································1371) 모형의 설명력 ·····················································································1372) 표준화 잔차의 공간적 자기상관 ·····················································1383) 모형의 평균 오차율 ···········································································1395. 개발된 예측 모형의 적용 ·····················································1421) 범죄발생 위험지도 작성 ···································································142(1) 다중선형회귀 모형에 의한 범죄발생 위험지도 ······················ 142(2) 지리가중회귀 모형에 의한 범죄발생 위험지도 ······················ 144(3) 회귀모형 간 범죄발생 위험지도의 비교 ··································1452) 방어기제 변수 조절을 통한 시나리오별 예측 결과 ···················148(1) 다중선형회귀 모형에 의한 시나리오별 예측 결과 ················ 149(2) 지리가중회귀 모형에 의한 시나리오별 예측 결과 ················ 1516. 소결 ···························································································156제6장 결론 ·····················································································1591. 연구의 결론 ·············································································1612. 연구의 한계점 및 향후 과제 ···············································164□ 참 고 문 헌 ·············································································166□ Abstract ···················································································174□ 감 사 의 글
최근 본 자료
댓글(0)
0