The primary aim of this dissertation was to develop and validate a methodology for identifying spatial clusters (hotspots) of various paediatric cancers within South Africa by using GIS software. The Hotspot Analysis (Getis-Ord Gi*) Tool was used for this purpose. A series of spatial clusters (hotspots) were identified by the tool for each cancer type and these clusters were compared with the exiting literature regarding known environmental and other carcinogens. The quality of the cancer data used in the dissertation was however found to be questionable and significantly underreported. This caused the results of the tool to also be questionable. The dissertation therefore concluded that the tool could be successfully used to identify spatial clusters of cancer in principle. It was however found that the results of the tool needed to be viewed without caution in this dissertation due to the low quality of the cancer data used. / MSc (Geography and Environmental Management), North-West University, Potchefstroom Campus, 2014

The primary aim of this dissertation was to develop and validate a methodology for identifying spatial clusters (hotspots) of various paediatric cancers within South Africa by using GIS software. The Hotspot Analysis (Getis-Ord Gi*) Tool was used for this purpose. A series of spatial clusters (hotspots) were identified by the tool for each cancer type and these clusters were compared with the exiting literature regarding known environmental and other carcinogens. The quality of the cancer data used in the dissertation was however found to be questionable and significantly underreported. This caused the results of the tool to also be questionable. The dissertation therefore concluded that the tool could be successfully used to identify spatial clusters of cancer in principle. It was however found that the results of the tool needed to be viewed without caution in this dissertation due to the low quality of the cancer data used. / MSc (Geography and Environmental Management), North-West University, Potchefstroom Campus, 2014

Tendo em vista o objetivo geral de analisar o perfil epidemiológico da epidemia de Aids, numa abordagem espaço-temporal dos casos de infecção por HIV e dos óbitos por Aids no município de São Paulo, de 1996 a 2007, o presente estudo usa um delineamento ecológico transversal. A unidade de análise é o Distrito Administrativo. A partir desta unidade são calculados os riscos relativos de se infectar por HIV e de ir a óbito por Aids no período definido. A análise temporal é feita a partir das situações dos anos 1996, 2000, 2004 e 2007, para cada gênero, que indicará a situação do momento e a tendência ao longo do tempo. A análise estatística espacial auxilia a localizar distritos de risco significativamente acima e/ou abaixo da média do município. Foram utilizados os dados referentes a óbitos por Aids por Distrito Administrativo Residencial do município de São Paulo para o período de 1996 a 2007, por gênero e faixa etária da OMS (Organização Mundial de Saúde), obtidos no SINAM do sítio do PRO-AIM (Programa de Aprimoramento da Informação de Mortalidade). Os dados referentes aos totais de população por distrito, ano, gênero e mesmas faixas etárias foram obtidos no sítio da Fundação SEADE. Para o ano 2000, os dados correspondem ao Censo do IBGE e para os demais, às estimativas populacionais desta Fundação. Correspondeu à base cartográfica digital do IBGE setores Censitários para o Município de São Paulo, ano 2000. Foram calculados os riscos relativos (RR) do período e para cada ano para cada distrito administrativo para a infecção por HIV e mortalidade por Aids, por gênero e total da população. O RR foi calculado utilizando-se o programa SaTScan. Os endereços foram geocodificados com o programa Maptitude, a partir da base de logradouros do município de São Paulo, utilizando-se o endereço (rua e número) e o CEP. Os agrupamentos espaciais foram mapeados usando-se a técnica corocromática. Os mapas de agrupamentos mostram os agrupamentos de risco alto em vermelho e de risco baixo, em azul. A evolução da epidemia teve grandes modificações, antes era uma doença que atingia a população de classe alta, hoje o panorama do HIV/Aids nos mostra que a população menos favorecida do ponto de vista socioeconomico é mais afetada devido a falta de informações e preparo para lidar com as medidas preventivas. Pode-se afirmar que a distribuição espaço-temporal da mortalidade por Aids e infecção por HIV caminham de maneira muito semelhante. Em ambos, ocorrem agrupamentos espaciais significativos na região central e casos abaixo do esperado nas regiões centro-sul, sudeste e sul. / Given the overall objective to analyze the epidemiological profile of the AIDS epidemic, in a space-time approach of the cases of HIV infection and AIDS deaths in São Paulo, from 1996 to 2007, this study uses an ecological design cross. The unit of analysis is the Administrative District. From this unit are calculated relative risks of becoming infected with HIV and of dying from AIDS within the defined period. The temporal analysis is done by the circumstances of the years 1996, 2000, 2004 and 2007 for each gender, which indicates the current situation and trends over time. The spatial statistical analysis helps districts find significantly higher risk and / or below average in the municipality. We used data on deaths from AIDS by Administrative District Residential in São Paulo for the period 1996 to 2007, by gender and age of the WHO (World Health Organization), obtained on SINAM Site PRO-AIM ( Improvement Program for Information Mortality). The data on population totals by district, year, genre and the same age groups were obtained at the site of SEADE. For 2000, data correspond to the IBGE Census and for others, the population estimation of the Foundation. It corresponded to the digital base map of IBGE - census tracts for the city of São Paulo, 2000. We calculated relative risks (RR) for the period and for each year in each administrative district to HIV infection and AIDS mortality by gender and total population. The RR was calculated using the program SaTScan. The addresses were geocoded using the Maptitude program, from the base to street addresses in the city of São Paulo, using the address (street and number) and zip code. The spatial clusters were mapped using the corocromática technique. The maps show high risk clusters in red and low risk, in blue. The epidemic has had major modifications before it was a disease that affects the upper class population, today the landscape of HIV / AIDS shows that the less favored population in terms of socioeconomic background is more affected due to the lack of information and preparation to deal with preventive measures. It can be argued that the spatial and temporal distribution of mortality from AIDS and HIV infection much like walking. In both, significant spatial clusters occur in the central site and lower than expected cases in south-central, southeast and south.

Made available in DSpace on 2017-07-10T16:32:28Z (GMT). No. of bitstreams: 1 Marines Rute De Oliveira Machado.pdf: 2572306 bytes, checksum: 2b3ca5a0ad44aa36c3bd64b4d5945716 (MD5) Previous issue date: 2014-12-11 / On this dissertation the economic development concept that was used takes place through the analysis of social and economic data, in which it is verified if the growth is related to the economic development of a place, coming from the principle that the development is due to the growth interconnected with the improvement in the quality of life of the population. Along with the new challenges that have been presented to the economy of the state of Paraná, it is essential to evaluate the technical, financial and political possibilities of reassuring the role of the local state as a correcting agent of interregional and social inequalities and to rethink the designs, the objectives and the orientation of social policies in order to make them innovative when facing the transformations that are occurring in society. Thus it is necessary to know how the economic development is distributed in the microregions of Cascavel, Toledo and Foz do Iguaçu, that form the western region of Paraná. In this sense the objective of this study was to analyze the economic development of municipalities of the microregions in the west of Paraná, verifying the determiners of the development level and classifying them according to the degree of economic development, conducting as well an exploratory analysis of spatial data. In using the technique of multivariate statistic, the factor analysis, an indicator was elaborated, the Index of Gross of Development to the 50 analyzed municipalities. It was verified, also, the spatial dependence of the calculated index. The results indicate, from the factor analysis applied to the model for the years of 2002 and 2012 the extraction of 5 factors that explain, jointly, 76,31% of total variance of the selected variables. In which the factor 1 is directly related to the quality of life of the population, factor 2 is related to urbanization and economic development, factor 3 is directly related to education, factor 4 is directly related to unemployment and income distribution and factor 5 is correlated to the GAV of services. The gross development index, which is the weighted average of each one of the 5 found factors to the years of 2002 and 2012, was of 3.27 and 6.17, respectively. Along with the classification of the analyzed municipalities it was possible to observe that only three municipalities were able to be classified with high development in 2002 and in 2012. Only 9 municipalities in 2002 achieved an average classification and, in 2012, this number increased to 19. The majority of municipalities got a classification of development considered to be low in 2002 (39) and, in 2012, this number stood at just (28). It was verified that the Index of Economic Development does not suffer influence of the spatial proximity, as the estimates that detect the spatial autocorrelation were not statistically significant. It was noted that there were no differences, in statistical terms, of the IBD in the majority of the studied municipalities, where through the cartograms few outliers were detected. As for the cluster formation, it was observed that the majority of municipalities have IBD that does not statistically differ from the average of the IBD from the western region. These results let it show that there is a homogeneity level in the region, not confirming the hypothesis of the research that the economic development was occurring in a different way among its municipalities. / Nesta dissertação, o conceito utilizado de desenvolvimento econômico se dá através da análise de dados sociais e econômicos, onde é verificado se o crescimento está relacionado ao desenvolvimento econômico de um local, partindo do princípio que o desenvolvimento é resultado do crescimento interligado com a melhora na qualidade de vida da população. Com os novos desafios que vem se apresentando para a economia paranaense, torna-se fundamental avaliar as possibilidades técnicas, financeiras e políticas de reafirmar o papel do Estado local como agente de correção das desigualdades interregionais e sociais e repensar os desenhos, os objetivos e a orientação de políticas sociais de forma a torná-las inovadoras diante das transformações que estão ocorrendo na sociedade. Assim, se faz necessário conhecer como está distribuído o desenvolvimento econômico das microrregiões de Cascavel, Toledo e Foz do Iguaçu, que formam a região oeste do Paraná. Neste sentido, o objetivo do presente estudo foi analisar o desenvolvimento econômico dos municípios das microrregiões do oeste do Paraná, verificando os fatores determinantes do nível de desenvolvimento e classificando-os quanto ao grau de desenvolvimento econômico, realizando, também, uma análise exploratória de dados espaciais. Usando a técnica da estatística multivariada, a análise fatorial, elaborou-se um indicador, o Índice de Bruto de Desenvolvimento para os 50 municípios analisados. Verificou-se, também, a dependência espacial do índice calculado. Os resultados apontam, a partir da análise fatorial aplicada ao modelo para os anos de 2002 e 2012 a extração de 5 fatores que explicam, em conjunto, 76,31% da variância total das variáveis selecionadas. Onde o fator 1 está diretamente relacionado com a qualidade de vida da população, o fator 2 relaciona-se com a urbanização e desempenho econômico, o fator 3 está diretamente relacionado com a educação, o fator 4 com o desemprego e distribuição de renda e o fator 5 apresenta relação com o Valor Agregado Bruto dos serviços. O índice de desenvolvimento bruto, que é a média ponderada de cada um dos 5 fatores encontrados para os anos de 2002 e 2012, foi de-3,27 e 6,17, respectivamente. Com a classificação dos municípios analisados foi possível observar que apenas três conseguiram ser classificados com desenvolvimento alto, em 2002 e em 2012. Apenas 9 municípios em 2002 conseguiram classificação média e, em 2012, esse número subiu para 19. A maioria dos municípios ficaram com classificação de desenvolvimento considerado baixo em 2002 (39) e, em 2012, esse número ficou em apenas (28). Verificou-se que o Índice de Desenvolvimento Econômico não sofre influência da proximidade espacial, uma vez que as estimativas que detectam a autocorrelação espacial não se mostraram estatisticamente significativas. Notou-se que não houve diferenças, em termos estatísticos, do IBD na maioria dos municípios estudados, onde através dos cartogramas detectou-se poucos outliers. Quanto a formação de clusters observou-se que a maioria dos municípios possuem IBD que não difere estatisticamente da média do IBD da região oeste. Estes resultados deixam transparecer que há na região um nível de homogeneidade, não confirmando a hipótese da pesquisa de que o desenvolvimento econômico estava ocorrendo de forma diferente entre seus municípios.

Tendo em vista o objetivo geral de analisar o perfil epidemiológico da epidemia de Aids, numa abordagem espaço-temporal dos casos de infecção por HIV e dos óbitos por Aids no município de São Paulo, de 1996 a 2007, o presente estudo usa um delineamento ecológico transversal. A unidade de análise é o Distrito Administrativo. A partir desta unidade são calculados os riscos relativos de se infectar por HIV e de ir a óbito por Aids no período definido. A análise temporal é feita a partir das situações dos anos 1996, 2000, 2004 e 2007, para cada gênero, que indicará a situação do momento e a tendência ao longo do tempo. A análise estatística espacial auxilia a localizar distritos de risco significativamente acima e/ou abaixo da média do município. Foram utilizados os dados referentes a óbitos por Aids por Distrito Administrativo Residencial do município de São Paulo para o período de 1996 a 2007, por gênero e faixa etária da OMS (Organização Mundial de Saúde), obtidos no SINAM do sítio do PRO-AIM (Programa de Aprimoramento da Informação de Mortalidade). Os dados referentes aos totais de população por distrito, ano, gênero e mesmas faixas etárias foram obtidos no sítio da Fundação SEADE. Para o ano 2000, os dados correspondem ao Censo do IBGE e para os demais, às estimativas populacionais desta Fundação. Correspondeu à base cartográfica digital do IBGE setores Censitários para o Município de São Paulo, ano 2000. Foram calculados os riscos relativos (RR) do período e para cada ano para cada distrito administrativo para a infecção por HIV e mortalidade por Aids, por gênero e total da população. O RR foi calculado utilizando-se o programa SaTScan. Os endereços foram geocodificados com o programa Maptitude, a partir da base de logradouros do município de São Paulo, utilizando-se o endereço (rua e número) e o CEP. Os agrupamentos espaciais foram mapeados usando-se a técnica corocromática. Os mapas de agrupamentos mostram os agrupamentos de risco alto em vermelho e de risco baixo, em azul. A evolução da epidemia teve grandes modificações, antes era uma doença que atingia a população de classe alta, hoje o panorama do HIV/Aids nos mostra que a população menos favorecida do ponto de vista socioeconomico é mais afetada devido a falta de informações e preparo para lidar com as medidas preventivas. Pode-se afirmar que a distribuição espaço-temporal da mortalidade por Aids e infecção por HIV caminham de maneira muito semelhante. Em ambos, ocorrem agrupamentos espaciais significativos na região central e casos abaixo do esperado nas regiões centro-sul, sudeste e sul. / Given the overall objective to analyze the epidemiological profile of the AIDS epidemic, in a space-time approach of the cases of HIV infection and AIDS deaths in São Paulo, from 1996 to 2007, this study uses an ecological design cross. The unit of analysis is the Administrative District. From this unit are calculated relative risks of becoming infected with HIV and of dying from AIDS within the defined period. The temporal analysis is done by the circumstances of the years 1996, 2000, 2004 and 2007 for each gender, which indicates the current situation and trends over time. The spatial statistical analysis helps districts find significantly higher risk and / or below average in the municipality. We used data on deaths from AIDS by Administrative District Residential in São Paulo for the period 1996 to 2007, by gender and age of the WHO (World Health Organization), obtained on SINAM Site PRO-AIM ( Improvement Program for Information Mortality). The data on population totals by district, year, genre and the same age groups were obtained at the site of SEADE. For 2000, data correspond to the IBGE Census and for others, the population estimation of the Foundation. It corresponded to the digital base map of IBGE - census tracts for the city of São Paulo, 2000. We calculated relative risks (RR) for the period and for each year in each administrative district to HIV infection and AIDS mortality by gender and total population. The RR was calculated using the program SaTScan. The addresses were geocoded using the Maptitude program, from the base to street addresses in the city of São Paulo, using the address (street and number) and zip code. The spatial clusters were mapped using the corocromática technique. The maps show high risk clusters in red and low risk, in blue. The epidemic has had major modifications before it was a disease that affects the upper class population, today the landscape of HIV / AIDS shows that the less favored population in terms of socioeconomic background is more affected due to the lack of information and preparation to deal with preventive measures. It can be argued that the spatial and temporal distribution of mortality from AIDS and HIV infection much like walking. In both, significant spatial clusters occur in the central site and lower than expected cases in south-central, southeast and south.

<p> This research examined the geographic variability as well as the individual-and neighbourhood-level determinants of overweight and obesity in Canada. Overweight and obesity represent a significant public health problem with grave implications for individuals as well as populations. Over the past two decades, the prevalence of overweight and obesity has reached epidemic proportions with the most substantial increases observed in economically developed countries. The World Health Organization indicated that globally 1.6 billion adults (age 15+) are overweight and at least 400 million adults were obese. In a Canadian context, recent data from Statistics Canada confirms that over the past twenty-five years, adult obesity rates in Canada have doubled (23% ), while childhood obesity rates have nearly tripled. </p> <p> Until recently, research has focused on biological and behavioural determinants of obesity, and currently there is a great deal of knowledge regarding the relationships between weight status and various risk factors at the individual-level (e.g. age, sex, socioeconomic deprivation, diet, physical activity). However, the majority of existing research has ignored the potential role played by the environment in the development of these conditions, despite a growing consensus that environmental and/or societal constraints may be major influences on increasing prevalence rates. </p> <p> Using data from the Canadian Community Health Surveys and the Desktop Mapping Information Technologies Incorporated spatial database, this research addressed the following objectives: (I) to examine sex-specific spatial patterns of overweight/obesity in Canada as well as investigate the presence of spatial clusters (2) to investigate the prevalence and determinants of overweight and obesity in Canada using spatial analysis and geographical information systems (GIS) and (3) to identify heterogeneities associated with the relationships between individual and socioenvironmental determinants and overweight and obesity at the individual-and community-levels. </p> <p> Results revealed marked geographical variation in overweight/obesity prevalence with higher values in the Northern and Atlantic health-regions and lower values in the Southern and Western health-regions of Canada. Significant positive spatial autocorrelation was found for both males and females, with significant clusters of high values or 'hot spots' of obesity in the Atlantic and Northern health-regions of Alberta, Saskatchewan, Manitoba and Ontario. Results also demonstrate the important role of the built-environment after adjustment demographic, socio-economic and behavioural characteristics. With regard to the built environment measures, landuse mix and residential density were found to be significantly associated with BMI. This study also demonstrated significant differences at the area-level of analysis, supporting related research that has suggested that individual-level factors alone cannot explain variation in obesity rates across space. In particular, average dwelling value was related to BMI independently of individual-level characteristics. Ultimately, this research has demonstrated that Canadian urban environments play a small but significant role in shaping the distribution of BMI. Yet, reversing current trends will require a multifaceted public health approach where interventions are developed from the individual-to the neighbourhood-level, specifically focusing on altering obesogenic environments. </p> / Thesis / Doctor of Philosophy (PhD)

Due to the advances in location-acquisition techniques, citizen geospatial data has emerged with opportunity for research, development, innovation, and business. A variety of research has been developed to study society and citizens through exploring patterns from geospatial data. In this thesis, we investigate patterns of population and human sentiments using GPS trajectory data and geo-tagged tweets. Kernel density estimation and emerging hot spot analysis are first used to demonstrate population distribution across space and time. Then a flow extraction model is proposed based on density difference for human movement detection and visualization. Case studies with volleyball game in West Lafayette and traffics in Puerto Rico verify the effectiveness of this method. Flow maps are capable of tracking clustering behaviors and direction maps drawn upon the orientation of vectors can precisely identify location of events. This thesis also analyzes patterns of human sentiments. Polarity of tweets is represented by a numeric value based on linguistics rules. Sentiments of four US college cities are analyzed according to its distribution on citizen, time, and space. The research result suggests that social media can be used to understand patterns of public sentiment and well-being.

Instituto de Pesquisa Econômica Aplicada / Because of the historical economic and social inequality, especially in rural areas the State of Bahia, and public policies for rural territorial development instituted in the 2000s, the general objective of this work is \"Discuss the borderline of Bahian rural areas, from the identification municipal types\". So initially it describes the evolution and characteristics of initiatives, policies and international and national public programs for rural and regional development that inspired the planning and actions of the Bahian government. Then presents the historical context and the specificities of inducing the development policies in Bahia, talking about the recent process of institution of Rural Territories programs (TRs), Identity Territories (IT) and Citizenship Territories (CTs). In support of scientific systematization, this work has a theoretical review of the development, especially in its rural and territorial aspect, and an empirical review, which presents the main international and national scientific papers, which assessed the development through multidimensional variables and by multivariate methods. All this depended on bibliographical and documentary research. The Bahia characteristics were analyzed, whose descriptive analysis and factor analysis (FA) recorded a state with a rural profile, but with economic, social and environmental conditions short of Brazilian reality. As a reference of the largest and lowest level of Municipal Human Development (HDI) territorial respectively, the Territory South Coast and the Sisal Territory have become objects of research. In order to describe in what aspects Bahian rural areas or are not similar to each other; and check whether they are made of homogeneous municipalities, a descriptive and comparative analysis was performed, and estimate a Rural Development Index (RDI). The results show that both territories record high number of municipalities, area and resident population, aspects that hinder the historical cohesion, cultural, social and economic. Although the two territories are consistent with the requirements of TRs programs, TCs and IT, the Territory Sisal has more rural profile and economic and social conditions most critical. The IDR shows heterogeneity in the two territories, although the Territory of Sisal, focus greater number of municipalities with lower levels of rural development. With the intention of \"propose methodologies for the delimitation of rural areas\", the results point 2 or 3 types of Spatial Clusters for both the Territory South Coast and for the Territory Sisal, groups composed of surrounding municipalities and with homogeneous characteristics. Through the main crops, the effective livestock, the various activities and displacement flow from the occupied have been identified different along guide ideas of each territory. Therefore, the results reveal that the South Coast Territory and the Sisal Territory are formed by heterogeneous municipalities; and that depending on the study variable and the type of method chosen, different possibilities for subdivision are possible territories. The important thing is to identify the demographic, social, cultural, economic, environmental characteristics; identify problems, challenges, possibilities, representative identity of territorial cohesion, able to promote rural development. / Por causa da histórica desigualdade econômica e social, principalmente no meio rural do estado da Bahia, e das políticas públicas de desenvolvimento territorial rural instituídas nos anos 2000, o objetivo geral desse trabalho é Discutir a delimitação dos territórios rurais baianos, a partir da identificação das tipologias municipais . Por isso, inicialmente descreve a evolução e as características das iniciativas, políticas e programas públicos internacionais e nacionais de desenvolvimento rural e territorial que inspiraram o planejamento e as ações do governo baiano. Em seguida apresenta o contexto histórico e as especificidades das políticas públicas indutoras do desenvolvimento na Bahia, discorrendo sobre o recente processo de instituição dos programas Territórios Rurais (TRs), Territórios da Identidade (TIs) e Territórios da Cidadania (TCs). Para fundamentar a sistematização científica, esse trabalho conta com uma revisão teórica sobre o desenvolvimento, sobretudo em sua vertente rural e territorial, e uma revisão empírica, que apresenta os principais trabalhos científicos internacionais e nacionais, os quais avaliaram o desenvolvimento através de variáveis multidimensionais e por métodos multivariados. Tudo isso dependeu da pesquisa bibliográfica e documental. Foram analisadas as características da Bahia, cuja análise descritiva e a Análise Fatorial (AF) apontaram um estado com perfil rural, mas com condições econômicas, sociais e ambientais aquém da realidade brasileira. Como referência do maior e do menor nível de Desenvolvimento Humano Municipal (IDH-M) territorial, respectivamente, o Território Litoral Sul e o Território do Sisal passaram a ser objetos da pesquisa. No intuito de descrever em quais aspectos os territórios rurais baianos são ou não semelhantes entre si; e verificar se são ou não constituídos por municípios homogêneos, foi realizada uma análise descritiva e comparativa, além de estimar um Índice de Desenvolvimento Rural (IDR). Os resultados apontam que ambos territórios registram elevado número de municípios, área e população residente, aspectos que dificultam a coesão histórica, cultural, social e econômica. Embora os dois territórios estejam condizentes com as exigências dos programas TRs, TCs e TIs, o Território do Sisal possui perfil mais rural e condições econômicas e sociais mais críticas. O IDR aponta heterogeneidade nos dois territórios, embora o Território do Sisal, concentre maior número de municípios com níveis inferiores de desenvolvimento rural. Com a intenção de Propor metodologias para a delimitação de territórios rurais , os resultados apontam 2 ou 3 tipos de Clusters Espaciais, tanto para o Território Litoral Sul como para o Território do Sisal, agrupamentos compostos por municípios circunvizinhos e com características homogêneas. Por meio das principais culturas, do efetivo da pecuária, das atividades diversas e do fluxo de deslocamento dos ocupados foram identificadas diferentes ideias-guia ao longo de cada território. Portanto, os resultados revelam que, o Território Litoral Sul e o Território do Sisal são formados por municípios heterogêneos; e, que dependendo da variável de estudo e do tipo de método escolhido, diferentes possibilidades de subdivisão para os territórios são possíveis. O importante é identificar as características demográficas, sociais, culturais, econômicas, ambientais; identificar os problemas, os desafios, as possibilidades, a identidade representativa da coesão territorial, capaz de promover o desenvolvimento rural. / Doutor em Economia

近來地理資訊系統(GIS)之資料庫受到不同領域的統計學家廣泛的研究，以期建立及分析可描述空間聚集效應及變異之模型，而描述空間相關存活資料之統計模式為公共衛生及流行病學上新興的研究議題。本文擬建立多維度半參數的貝氏階層模型，並結合空間及非空間隨機效應以描述存活資料中的空間變異。此模式將利用多變量條件自回歸(MCAR)模型以檢驗在不同地理區域中是否存有空間聚集效應。而基準風險函數之生成為分析貝氏半參數階層模型的重要步驟，本研究將利用混合Polya樹之方式生成基準風險函數。美國國家癌症研究院之「流行病監測及最終結果」(Surveillance Epidemiology and End Results, SEER)資料庫為目前美國最完整的癌症病人長期追蹤資料，包含癌症病人存活狀況、多重癌症史、居住地區及其他分析所需之個人資料。本文將自此資料庫擷取美國愛荷華州之癌症病人資料為例作實證分析，並以貝氏統計分析中常用之模型比較標準如條件預測指標(CPO)、平均對數擬邊際概似函數值(ALMPL)、離差訊息準則(DIC)分別測試其可靠度。 / The databases of Geographic Information System (GIS) have gained attention among different fields of statisticians to develop and analyze models which account for spatial clustering and variation. There is an emerging interest in modeling spatially correlated survival data in public health and epidemiologic studies. In this article, we develop Bayesian multivariate semiparametric hierarchical models to incorporate both spatially correlated and uncorrelated frailties to answer the question of spatial variation in the survival patterns, and we use multivariate conditionally autoregressive (MCAR) model to detect that whether there exists the spatial cluster across different areas. The baseline hazard function will be modeled semiparametrically using mixtures of finite Polya trees. The SEER (Surveillance Epidemiology and End Results) database from the National Cancer Institute (NCI) provides comprehensive cancer data about patient’s survival time, regional information, and others demographic information. We implement our Bayesian hierarchical spatial models on Iowa cancer data extracted from SEER database. We illustrate how to compute the conditional predictive ordinate (CPO), the average log-marginal pseudo-likelihood (ALMPL), and deviance information criterion (DIC), which are Bayesian criterions for model checking and comparison among competing models.