abstract: The volume of available spatial data has increased tremendously. Such data includes but is not limited to: weather maps, socioeconomic data, vegetation indices, geotagged social media, and more. These applications need a powerful data management platform to support scalable and interactive analytics on big spatial data. Even though existing single-node spatial database systems (DBMSs) provide support for spatial data, they suffer from performance issues when dealing with big spatial data. Challenges to building large-scale spatial data systems are as follows: (1) System Scalability: The massive-scale of available spatial data hinders making sense of it using traditional spatial database management systems. Moreover, large-scale spatial data, besides its tremendous storage footprint, may be extremely difficult to manage and maintain due to the heterogeneous shapes, skewed data distribution and complex spatial relationship. (2) Fast analytics: When the user runs spatial data analytics applications using graphical analytics tools, she does not tolerate delays introduced by the underlying spatial database system. Instead, the user needs to see useful information quickly. In this dissertation, I focus on designing efficient data systems and data indexing mechanisms to bolster scalable and interactive analytics on large-scale geospatial data. I first propose a cluster computing system GeoSpark which extends the core engine of Apache Spark and Spark SQL to support spatial data types, indexes, and geometrical operations at scale. In order to reduce the indexing overhead, I propose Hippo, a fast, yet scalable, sparse database indexing approach. In contrast to existing tree index structures, Hippo stores disk page ranges (each works as a pointer of one or many pages) instead of tuple pointers in the indexed table to reduce the storage space occupied by the index. Moreover, I present Tabula, a middleware framework that sits between a SQL data system and a spatial visualization dashboard to make the user experience with the dashboard more seamless and interactive. Tabula adopts a materialized sampling cube approach, which pre-materializes samples, not for the entire table as in the SampleFirst approach, but for the results of potentially unforeseen queries (represented by an OLAP cube cell). / Dissertation/Thesis / Doctoral Dissertation Computer Science 2020

A tremendous amount of information is being shared every day on social media sites such as Facebook, Twitter or Google+. However, only a small portion of users provide their location information, which can be helpful in targeted advertising and many other services. Current methods in location estimation using social relationships consider social friendship as a simple binary relationship. However, social closeness between users and structure of friends have strong implications on geographic distances. In the first task, we introduce new measures to evaluate the social closeness between users and structure of friends. Then we propose models that use them for location estimation. Compared with the models which take the friend relation as a binary feature, social closeness can help identify which friend of a user is more important and friend structure can help to determine significance level of locations, thus improving the accuracy of the location estimation models. A confidence iteration method is further introduced to improve estimation accuracy and overcome the problem of scarce location information. We evaluate our methods on two different datasets, Twitter and Gowalla. The results show that our model can improve the estimation accuracy by 5% - 20% compared with state-of-the-art friend-based models. In the second task, we also propose a Local Event Discovery and Summarization (LEDS) framework to detect local events from Twitter. Many existing algorithms for event detection focus on larger-scale events and are not sensitive to smaller-scale local events. Most of the local events detected by these methods are major events like important sports, shows, or big natural disasters. In this work, we propose the LEDS framework to detect both bigger and smaller events. LEDS contains three key steps: 1) Detecting possible event related terms by monitoring abnormal distribution in different locations and times; 2) Clustering tweets based on their key terms, time, and location distribution; and 3) Extracting descriptions include time, location, and key sentences of local events from clusters. The model is evaluated on a real-world Twitter dataset with more than 60 million tweets. The analysis of Twitter data can help to predict or explain many real-world phenomena. The relationships among events in the real world can be reflected among the topics on social media. In the third task, we propose the concept of topic association and the associated mining algorithms. Topics with close temporal and spatial relationship may have direct or potential association in the real world. Our goal is to mine such topic associations and show their relationships in different time-region frames. We propose to use the concepts of participation ratio and participation index to measure the closeness among topics and propose a spatiotemporal index to calculate them efficiently. With the topic filtering and the topic combination, we further optimize the mining process and the mining results.

No description available.

No description available.

Creating physical simulations between virtual and real objects in Augmented Reality (AR) is essential for the user experience. Otherwise the user might lose sense of depth, distance and size. One of these simulations is occlusion, meaning that virtual content should be partially or fully occluded if real world objects is in the line-of-sight between the user and the content. The challenge for simulating occlusion is to construct the geometric model of the current AR environment. Earlier studies within the field have all tried to create realistic pixel-perfect occlusion and most of them have either required depth-sensing hardware or a static predefined environment. This study proposes and evaluates an alternative model-based approach to the problem. It uses geospatial data to construct the geometric model of all the buildings in the current environment, making virtual content occluded by all real buildings in the current environment. This approach made the developed function compatible with non depth-sensing devices and in a dynamic outdoor urban environment. To evaluate the solution it was implemented in a sensor-based AR application visualizing a future building in Stockholm. The effect of the developed function was that the future virtual building was occluded as expected. However, it was not pixel-perfect, meaning that the simulated occlusion was not realistic, but results from the conducted user study said that it fulfilled its goal. A majority of the participants thought that their AR experience got better with the solution activated and that their depth perception improved. However, any definite conclusions could not be drawn due to issues with the sensor-based tracking. The result from this study is interesting for the mobile AR field since the great majority of smartphones are not equipped with depth sensors. Using geospatial data for simulating occlusions, or other physical interactions between virtual and real objects, could then be an efficient enough solution until depth-sensing AR devices are more widely used. / För att uppnå en god användarupplevelse i Augmented Reality (AR) så är det viktigt att simulera fysiska interaktioner mellan de virtuella och reella objekten. Om man inte gör det kan användare uppfatta saker som djup, avstånd och storlek felaktigt. En av dessa simulationer är ocklusion som innebär att det virtuella innehållet ska vara delvis eller helt ockluderat om ett reellt objekt finns i siktlinjen mellan användaren och innehållet. För att simulera detta är utmaningen att rekonstruera den geometriska modellen av den nuvarande miljön.Tidigare studier inom fältet har försökt att uppnå en perfekt simulation av ocklusion, men majoriteten av dem har då krävt antingen djupavkännande hårdvara eller en statisk fördefinierad miljö. Denna studie föreslår och utvärderar en alternativ modellbaserad lösning på problemet. Lösningen använder geospatial data för att rekonstruera den geometriska modellen av alla byggnader i den nuvarande omgivningen, vilket resulterar i att det virtuella innehållet blir ockluderat av alla reella byggnader i den nuvarande miljön. Den utvecklade funktionen blev i och med det kompatibel på icke djupavkännande enheter och fungerande i en dynamisk urban miljö. För attutvärdera denna funktion så var den implementerad i en sensorbaserad AR applikation som visualiserade en framtida byggnad i Stockholm. Resultatet visade att den utvecklade funktionen ockluderade den virtuella byggnaden som förväntat. Dock gjorde den ej det helt realistiskt, men resultatet från den utförda användarstudien visade att den uppnådde sitt mål. Majoriteten av deltagarna ansåg att deras AR upplevelse blev bättre med den utvecklade funktionen aktiverad och ett deras uppfattning av djup förbättrades. Dock kan inga definitiva slutsatser dras eftersom AR applikationen hade problem med den sensorbaserade spårningen. Resultaten är intressant för det mobila AR fältet eftersom majoriteten av alla smartphones ej har stöd för djupavkänning. Att använda geospatial data för att simulera ocklusion, eller någon annan fysisk interaktion mellan virtuella och reella objekt, kan då vara en tillräckligt effektiv lösning tills djupavkännande AR enheter används mer.

Submitted by Eunice Novais (enovais@uepg.br) on 2017-09-06T23:22:17Z No. of bitstreams: 2 license_rdf: 811 bytes, checksum: e39d27027a6cc9cb039ad269a5db8e34 (MD5) Silvia Mantuani.pdf: 1063712 bytes, checksum: fb17dc1ed6cea5a6dc334c161364c066 (MD5) / Made available in DSpace on 2017-09-06T23:22:17Z (GMT). No. of bitstreams: 2 license_rdf: 811 bytes, checksum: e39d27027a6cc9cb039ad269a5db8e34 (MD5) Silvia Mantuani.pdf: 1063712 bytes, checksum: fb17dc1ed6cea5a6dc334c161364c066 (MD5) Previous issue date: 2017-07-10 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / A segurança, qualidade e a origem dos alimentos são focos dos consumidores atuais, que buscam informações relacionadas ao sistema de produção e cuidados com o local de produção. Sistemas de rastreabilidade convencionais correspondem a uma tecnologia adequada para analisar informações do produto em qualquer etapa na cadeia produtiva, porém não disponibilizam os dados específicos em relação ao local e ao entorno onde os produtos ou lotes de produtos foram produzidos. A geo-rastreabilidade permite complementar essa carência dos sistemas de rastreabilidade, possibilitando abranger informações geográficas sobre o produto. A associação de indicadores geográficos e demais informações resulta na melhoria da segurança do produto rastreado. O RastroGrão é um framework de rastreabilidade de grãos que registra dados dos agentes da cadeia de produção para posterior consulta pelo consumidor final, porém não foi modelado para disponibilizar a geoinformação. O objetivo desta dissertação é apresentar a especificação da integração de dados geoespaciais para o RastroGrão. Essa pesquisa foi baseada nos regulamentos e normas para integração de dados geoespaciais, além de análise de sistemas de gestão de geoinformação e análise nos softwares web GeoTraceAgri, GTIS CAP, GeoFairTrade, GeoWine e GeoRastro, que implementam geoinformação integrada a dados de rastreabilidade de cadeias produtivas. Para integrar as informações geográficas existentes foi necessário utilizar a Infraestrutura de Dados Espaciais (IDE), para combinar diversas fontes de dados, originando informações sobre a área analisada. Com a integração dos dados geoespaciais, os agentes da cadeia produtiva e os consumidores têm informações precisas sobre os produtos que consomem, com verificação do local e entorno, onde foi produzido, transportado e armazenado o produto, além das práticas envolvidas na produção de determinado produto. Esta integração, com indicadores, auxilia na garantia da segurança do produto e proteção do ambiente, além de proporcionar controle agrícola sustentável. Palavras-chave: Geoinformação, Geo-Rastreabilidade, Dados Geoespaciais, Integração. / The safety, quality and origin of food are the focus of current consumers, who seek information related to the system of production and care with the place of production. Conventional traceability systems correspond to appropriate technology for analyzing product information at any stage in the production chain, but they do not make specific data available in relation to the location and environment in which the products or batches of products were produced. Geotraceability allows us to complement this lack of traceability systems, making it possible to cover geographic information about the product. The association of geographic indicators and other information results in improved security of the traced product. RastroGrão is a grain traceability framework that records data from the agents of the production chain for later consultation by the final consumer, but was not modeled to make geoinformation available. The objective of this dissertation is to present the specification of geospatial data integration for RastroGrão. This research was based on the regulations and standards for the integration of geospatial data, as well as analysis of geoinformation management systems and analysis in GeoTraceAgri, GTIS CAP, GeoFairTrade, GeoWine and GeoRastro web software, which implement integrated geoinformation to traceability data of productive chains. In order to integrate the existing geographic information, it was necessary to use the Spatial Data Infrastructure (SDI) to combine several data sources, giving information about the analyzed area. With the integration of geospatial data, the agents of the production chain and consumers have accurate information about the products they consume, with verification of the location and environment, where the product was produced, transported and stored, as well as the practices involved in the production of a given product. This integration, with indicators, assists in ensuring product safety and protection of the environment, as well as providing sustainable agricultural control.

Research Report submitted in partial fulfilment for the degree of Master of Science (Geographical Information Systems and Remote Sensing) School of Geography, Archaeology and Environmental Studies, University of the Witwatersrand, Johannesburg. August 2016. / The thematic classification of land use and land cover (LULC) from remotely sensed imagery data is one of the most common research branches of applied remote sensing sciences. The performances of the pixel-based image analysis (PBIA) and object-based image analysis (OBIA) Support Vector Machine (SVM) learning algorithms were subjected to comparative assessment using WorldView-2 and SPOT-6 multispectral images of the Mokhotlong District in Lesotho covering approximately an area of 100 km2. For this purpose, four LULC classification models were developed using the combination of SVM –based image analysis approach (i.e. OBIA and/or PBIA) on high resolution images (WorldView-2 and/or SPOT-6) and the results were subjected to comparisons with one another. Of the four LULC models, the OBIA and WorldView-2 model (overall accuracy 93.2%) was found to be more appropriate and reliable for remote sensing application purposes in this environment. The OBIA-WorldView-2 LULC model was subjected to spatial overlay analysis with DEM derived topographic variables in order to evaluate the relationship between the spatial distribution of LULC types and topography, particularly for topographically-controlled patterns. It was discovered that although that there are traces of the relationship between the LULC types distributions and topography, it was significantly convoluted due to both natural and anthropogenic forces such that the topographic-induced patterns for most of the LULC types had been substantial disrupted. / LG2017

Three-dimensional laser scanning has revolutionized spatial data acquisition and can be completed from a variety of platforms including airborne (ALS), mobile (MLS), and static terrestrial (TLS) laser scanning. MLS is a rapidly evolving technology that provides increases in efficiency and safety over static TLS, while still providing similar levels of accuracy and resolution. The componentry that make up a MLS system are more parallel to Airborne Laser Scanning (ALS) than to that of TLS. However, achievable accuracies, precisions, and resolution results are not clearly defined for MLS systems. As such, industry professionals need guidelines to standardize the process of data collection, processing, and reporting. This thesis lays the foundation for MLS guidelines with a thorough review of currently available literature that has been completed in order to demonstrate the capabilities and limitations of a generic MLS system. A key difference between MLS and TLS is that a mobile platform is able to collect a continuous path of geo-referenced points along the navigation path, while a TLS collects points from many separate reference frames as the scanner is moved from location to location. Each individual TLS setup must be registered (linked with a common coordinate system) to adjoining scan setups. A study was completed comparing common methods of TLS registration and geo-referencing (e.g., target, cloud-cloud, and hybrid methods) to assist a TLS surveyor in deciding the most appropriate method for their projects. Results provide insight into the level of accuracy (mm to cm level) that can be achieved using the various methods as well as the field collection and office processing time required to obtain a fully geo-referenced point cloud. Lastly, a quality assurance methodology has been developed for any form of LiDAR data to verify both the absolute and relative accuracy of a point cloud without the use of retro-reflective targets. This methodology incorporates total station validation of a scanners point cloud to compare slopes of common features. The comparison of 2D slope features across a complex geometry of cross-sections provides 3D positional error in both horizontal and vertical component. This methodology lowers the uncertainty of single point accuracy statistics for point clouds by utilizing a larger portion of a point cloud for statistical accuracy verification. This use of physical features for accuracy validation is particularly important for MLS systems because MLS systems cannot produce sufficient resolution on targets for accuracy validation unless they are placed close to the vehicle. / Graduation date: 2012

No description available.

Analytical problems caused by over-fitting, confounding and non-independence in the data is a major challenge for variable selection. As more variables are tested against a certain data set, there is a greater risk that some will explain the data merely by chance, but will fail to explain new data. The main aim of this study is to employ a systematic and practicable variable selection process for the spatial analysis and mapping of historical malaria risk in Botswana using data collected from the MARA (Mapping Malaria Risk in Africa) project and environmental and climatic datasets from various sources. Details of how a spatial database is compiled for a statistical analysis to proceed is provided. The automation of the entire process is also explored. The final bayesian spatial model derived from the non-spatial variable selection procedure using Markov Chain Monte Carlo simulation was fitted to the data. Winter temperature had the greatest effect of malaria prevalence in Botswana. Summer rainfall, maximum temperature of the warmest month, annual range of temperature, altitude and distance to closest water source were also significantly associated with malaria prevalence in the final spatial model after accounting for spatial correlation. Using this spatial model malaria prevalence at unobserved locations was predicted, producing a smooth risk map covering Botswana. The automation of both compiling the spatial database and the variable selection procedure proved challenging and could only be achieved in parts of the process. The non-spatial selection procedure proved practical and was able to identify stable explanatory variables and provide an objective means for selecting one variable over another, however ultimately it was not entirely successful due to the fact that a unique set of spatial variables could not be selected.