Alexandre Sorokine, R. Karthik, A. King, B. Bhaduri
Big Data has already proven itself as a valuable tool that lets geographers and urban researchers utilize large data resources to generate new insights. However, wider adoption of Big Data techniques in these areas is impeded by a number of difficulties in both knowledge discovery and data and science production. Typically users face such problems as disparate and scattered data, data management, spatial searching, insufficient computational capacity for data-driven analysis and modelling, and the lack of tools to quickly visualize and summarize large data and analysis results. Here we propose an architecture for an Urban Information System (UrbIS) that mitigates these problems by utilizing the Big Data as a Service (BDaaS) concept. With technological roots in High-performance Computing (HPC), BDaaS is based on the idea of outsourcing computations to different computing paradigms, scalable to super-computers. UrbIS aims to incorporate federated metadata search, integrated modeling and analysis, and geovisualization into a single seamless workflow. The system is under active development and is built around various emerging technologies that include hybrid and NoSQL databases, massively parallel systems, GPU computing, and WebGL-based geographic visualization. UrbIS is designed to facilitate the use of Big Data across multiple cities to better understand how urban areas impact the environment and how climate change and other environmental change impact urban areas.
{"title":"Big data as a service from an urban information system","authors":"Alexandre Sorokine, R. Karthik, A. King, B. Bhaduri","doi":"10.1145/3006386.3006391","DOIUrl":"https://doi.org/10.1145/3006386.3006391","url":null,"abstract":"Big Data has already proven itself as a valuable tool that lets geographers and urban researchers utilize large data resources to generate new insights. However, wider adoption of Big Data techniques in these areas is impeded by a number of difficulties in both knowledge discovery and data and science production. Typically users face such problems as disparate and scattered data, data management, spatial searching, insufficient computational capacity for data-driven analysis and modelling, and the lack of tools to quickly visualize and summarize large data and analysis results. Here we propose an architecture for an Urban Information System (UrbIS) that mitigates these problems by utilizing the Big Data as a Service (BDaaS) concept. With technological roots in High-performance Computing (HPC), BDaaS is based on the idea of outsourcing computations to different computing paradigms, scalable to super-computers. UrbIS aims to incorporate federated metadata search, integrated modeling and analysis, and geovisualization into a single seamless workflow. The system is under active development and is built around various emerging technologies that include hybrid and NoSQL databases, massively parallel systems, GPU computing, and WebGL-based geographic visualization. UrbIS is designed to facilitate the use of Big Data across multiple cities to better understand how urban areas impact the environment and how climate change and other environmental change impact urban areas.","PeriodicalId":416086,"journal":{"name":"International Workshop on Analytics for Big Geospatial Data","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-10-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121582748","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
S. Migliorini, A. Belussi, Mauro Negri, G. Pelagatti
Spatial data usually encapsulate semantic characterization of features which carry out important meaning and relations among objects, such as the containment between the extension of a region and of its constituent parts. The GeoUML methodology allows one to bring the gap between the definition of spatial integrity constraints at conceptual level and the realization of validation procedures. In particular, it automatically generates SQL validation queries starting from a conceptual specification and using predefined SQL templates. These queries can be used to check data contained into spatial relational databases, such as PostGIS. However, the quality requirements and the amount of available data are considerably growing making unfeasible the execution of these validation procedures. The use of the map-reduce paradigm can be effectively applied in such context since the same test can be performed in parallel on different data chunks and then partial results can be combined together to obtain the final set of violating objects. Pigeon is a data-flow language defined on top of Spatial Hadoop which provides spatial data types and functions. The aim of this paper is to explore the possibility to extend the GeoUML methodology by automatically producing Pigeon validation procedures starting from a set of predefined Pigeon macros. These scripts can be used in a map-reduce environment in order to make feasible the validation of large datasets.
{"title":"Towards massive spatial data validation with SpatialHadoop","authors":"S. Migliorini, A. Belussi, Mauro Negri, G. Pelagatti","doi":"10.1145/3006386.3006392","DOIUrl":"https://doi.org/10.1145/3006386.3006392","url":null,"abstract":"Spatial data usually encapsulate semantic characterization of features which carry out important meaning and relations among objects, such as the containment between the extension of a region and of its constituent parts. The GeoUML methodology allows one to bring the gap between the definition of spatial integrity constraints at conceptual level and the realization of validation procedures. In particular, it automatically generates SQL validation queries starting from a conceptual specification and using predefined SQL templates. These queries can be used to check data contained into spatial relational databases, such as PostGIS.\u0000 However, the quality requirements and the amount of available data are considerably growing making unfeasible the execution of these validation procedures. The use of the map-reduce paradigm can be effectively applied in such context since the same test can be performed in parallel on different data chunks and then partial results can be combined together to obtain the final set of violating objects. Pigeon is a data-flow language defined on top of Spatial Hadoop which provides spatial data types and functions. The aim of this paper is to explore the possibility to extend the GeoUML methodology by automatically producing Pigeon validation procedures starting from a set of predefined Pigeon macros. These scripts can be used in a map-reduce environment in order to make feasible the validation of large datasets.","PeriodicalId":416086,"journal":{"name":"International Workshop on Analytics for Big Geospatial Data","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-10-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127913048","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The increasing pervasiveness of location-aware technologies is leading to the rise of large, spatio-temporal datasets and to the opportunity of discovering usable knowledge about the behaviors of people and objects. Applied extensively in transportation, spatial big data and its analytics can deliver useful insights on a number of different issues such as congestion, delays, public transport reliability and so on. Predominantly studied for its use in operational management, spatial big data can be used to provide insight in strategic applications as well, from planning and design to evaluation and management. Such large scale, streaming spatial big data can be used in the improvement of public transport, for example the design of public transport networks and reliability. In this paper, we analyze GTFS data from the cities of Stockholm and Rome to gain insight on the sources and factors influencing public transport delays in the cities. The analysis is performed on a combination of GTFS data with data from other sources. The paper points to key issues in the analysis of real time data, driven by the contextual setting in the two cities.
{"title":"Analytics on public transport delays with spatial big data","authors":"Jayanth Raghothama, V. M. Shreenath, S. Meijer","doi":"10.1145/3006386.3006387","DOIUrl":"https://doi.org/10.1145/3006386.3006387","url":null,"abstract":"The increasing pervasiveness of location-aware technologies is leading to the rise of large, spatio-temporal datasets and to the opportunity of discovering usable knowledge about the behaviors of people and objects. Applied extensively in transportation, spatial big data and its analytics can deliver useful insights on a number of different issues such as congestion, delays, public transport reliability and so on. Predominantly studied for its use in operational management, spatial big data can be used to provide insight in strategic applications as well, from planning and design to evaluation and management. Such large scale, streaming spatial big data can be used in the improvement of public transport, for example the design of public transport networks and reliability. In this paper, we analyze GTFS data from the cities of Stockholm and Rome to gain insight on the sources and factors influencing public transport delays in the cities. The analysis is performed on a combination of GTFS data with data from other sources. The paper points to key issues in the analysis of real time data, driven by the contextual setting in the two cities.","PeriodicalId":416086,"journal":{"name":"International Workshop on Analytics for Big Geospatial Data","volume":"58 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-10-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132526626","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
J. Heo, Sanghyun Yoon, Won Seob Oh, J. Ma, Sungha Ju, S. Yun
Spatial big data (SBD) has been utilized in many fields and we propose SBD analytics to apply to education with semantic trajectory data of undergraduate students in Songdo International Campus at Yonsei University. Higher education is under a pressure of disruptive innovation, so that colleges and universities strive to provide not only better education but also customized service to every single student, for a matter of survival in upcoming drastic wave. The entire research plan is to present a smart campus with SBD analytics for education, safety, health, and campus management, and this research is composed of four specific items: (1) to produce 3D mapping for project site; (2) to build semantic trajectory based on class attendance records, dorm gate entry records, etc.; (3) to collect pedagogical and other parameters of students; (4) to find relationship among trajectory patterns and pedagogical characteristics. Successful completion of the research would set a milestone to use semantic trajectory to predict student performance and characteristics, even further to go to proactive student care system and student activity guiding system. It can eventually present better customized education services to participating students.
{"title":"Spatial computing goes to education and beyond: can semantic trajectory characterize students?","authors":"J. Heo, Sanghyun Yoon, Won Seob Oh, J. Ma, Sungha Ju, S. Yun","doi":"10.1145/3006386.3006389","DOIUrl":"https://doi.org/10.1145/3006386.3006389","url":null,"abstract":"Spatial big data (SBD) has been utilized in many fields and we propose SBD analytics to apply to education with semantic trajectory data of undergraduate students in Songdo International Campus at Yonsei University. Higher education is under a pressure of disruptive innovation, so that colleges and universities strive to provide not only better education but also customized service to every single student, for a matter of survival in upcoming drastic wave. The entire research plan is to present a smart campus with SBD analytics for education, safety, health, and campus management, and this research is composed of four specific items: (1) to produce 3D mapping for project site; (2) to build semantic trajectory based on class attendance records, dorm gate entry records, etc.; (3) to collect pedagogical and other parameters of students; (4) to find relationship among trajectory patterns and pedagogical characteristics. Successful completion of the research would set a milestone to use semantic trajectory to predict student performance and characteristics, even further to go to proactive student care system and student activity guiding system. It can eventually present better customized education services to participating students.","PeriodicalId":416086,"journal":{"name":"International Workshop on Analytics for Big Geospatial Data","volume":"80 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-10-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127307234","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The rapid growing volumes of spatial data have brought significant challenges on developing high-performance spatial data processing techniques in parallel and distributed computing environments. Spatial joins are important data management techniques in gaining insights from large-scale geospatial data. While several distributed spatial join techniques based on spatial partitions have been implemented on top of existing Big Data systems, they are not capable of natively exploiting massively data parallel computing power provided by modern commodity Graphics Processing Units (GPUs). In this study, as an important component of our research initiative in developing high-performance spatial join techniques on GPUs, we have designed and implemented a polyline intersection based spatial join technique that is capable of exploiting massively data parallel computing power on GPUs. The proposed polyline intersection based spatial join technique is integrated into a customized lightweight distributed execution engine that natively supports spatial partitions. We empirically evaluate the performance of the proposed spatial join technique on both a standalone GPU-equipped workstation and Amazon EC2 GPU-accelerated clusters and demonstrate its high performance when comparing with the state-of-the-art.
{"title":"High-performance polyline intersection based spatial join on GPU-accelerated clusters","authors":"Simin You, Jianting Zhang, L. Gruenwald","doi":"10.1145/3006386.3006390","DOIUrl":"https://doi.org/10.1145/3006386.3006390","url":null,"abstract":"The rapid growing volumes of spatial data have brought significant challenges on developing high-performance spatial data processing techniques in parallel and distributed computing environments. Spatial joins are important data management techniques in gaining insights from large-scale geospatial data. While several distributed spatial join techniques based on spatial partitions have been implemented on top of existing Big Data systems, they are not capable of natively exploiting massively data parallel computing power provided by modern commodity Graphics Processing Units (GPUs). In this study, as an important component of our research initiative in developing high-performance spatial join techniques on GPUs, we have designed and implemented a polyline intersection based spatial join technique that is capable of exploiting massively data parallel computing power on GPUs. The proposed polyline intersection based spatial join technique is integrated into a customized lightweight distributed execution engine that natively supports spatial partitions. We empirically evaluate the performance of the proposed spatial join technique on both a standalone GPU-equipped workstation and Amazon EC2 GPU-accelerated clusters and demonstrate its high performance when comparing with the state-of-the-art.","PeriodicalId":416086,"journal":{"name":"International Workshop on Analytics for Big Geospatial Data","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-10-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132003670","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
A domain expert can process heterogeneous data to make meaningful interpretations or predictions from the data. For example, by looking at research papers and patent records, an expert can determine the maturity of an emerging technology and predict the geographic location(s) and time (e.g., in a certain year) where and when the technology will be a success. However, this is an expert- and manual-intensive task. This paper presents an end-to-end system that integrates heterogeneous data sources into a knowledge graph in the RDF (Resource Description Framework) format using an ontology. Then the user can easily query the knowledge graph to prepare the required data for different types of predictive analysis tools. We show a case study of predicting the (geographic) center(s) of fuel cell technologies using data collected from public sources to demonstrate the feasibility of our system. The system extracts, cleanses, and augments data from public sources including research papers and patent records. Next, the system uses an ontology-based data integration method to generate knowledge graphs in the RDF format to enable users to switch quickly between machine learning models for predictive analytic tasks. We tested the system using the Support Vector Machine and Multiple Hidden Markov Models and achieved 66.7% and 83.3% accuracy on the city and year levels of spatial and temporal resolutions, respectively.
{"title":"Building knowledge graph from public data for predictive analysis: a case study on predicting technology future in space and time","authors":"Weiwei Duan, Yao-Yi Chiang","doi":"10.1145/3006386.3006388","DOIUrl":"https://doi.org/10.1145/3006386.3006388","url":null,"abstract":"A domain expert can process heterogeneous data to make meaningful interpretations or predictions from the data. For example, by looking at research papers and patent records, an expert can determine the maturity of an emerging technology and predict the geographic location(s) and time (e.g., in a certain year) where and when the technology will be a success. However, this is an expert- and manual-intensive task. This paper presents an end-to-end system that integrates heterogeneous data sources into a knowledge graph in the RDF (Resource Description Framework) format using an ontology. Then the user can easily query the knowledge graph to prepare the required data for different types of predictive analysis tools. We show a case study of predicting the (geographic) center(s) of fuel cell technologies using data collected from public sources to demonstrate the feasibility of our system. The system extracts, cleanses, and augments data from public sources including research papers and patent records. Next, the system uses an ontology-based data integration method to generate knowledge graphs in the RDF format to enable users to switch quickly between machine learning models for predictive analytic tasks. We tested the system using the Support Vector Machine and Multiple Hidden Markov Models and achieved 66.7% and 83.3% accuracy on the city and year levels of spatial and temporal resolutions, respectively.","PeriodicalId":416086,"journal":{"name":"International Workshop on Analytics for Big Geospatial Data","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-10-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114461338","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Qiang Zhang, Ranga Raju Vatsavai, Ashwin Shashidharan, D. Berkel
The simulation of urban growth is an important part of urban planning and development. Due to large data and computational challenges, urban growth simulation models demand efficient data analytic frameworks for scaling them to large geographic regions. Agent-based models are widely used to observe and analyze the urban growth simulation at various scales. The incorporation of the agent-based model makes the scaling task even harder due to communication and coordination among agents. Many existing agent-based model frameworks were implemented using traditional shared and distributed memory programming models. On the other hand, Apache Spark is becoming a popular platform for distributed big data in-memory analytics. This paper presents an implementation of agent-based sub-model in Apache Spark framework. With the in-memory computation, Spark implementation outperforms the traditional distributed memory implementation using MPI. This paper provides (i) an overview of our framework capable of running urban growth simulations at a fine resolution of 30 meter grid cells, (ii) a scalable approach using Apache Spark to implement an agent-based model for simulating human decisions, and (iii) the comparative analysis of performance of Apache Spark and MPI based implementations.
{"title":"Agent based urban growth modeling framework on Apache Spark","authors":"Qiang Zhang, Ranga Raju Vatsavai, Ashwin Shashidharan, D. Berkel","doi":"10.1145/3006386.3007610","DOIUrl":"https://doi.org/10.1145/3006386.3007610","url":null,"abstract":"The simulation of urban growth is an important part of urban planning and development. Due to large data and computational challenges, urban growth simulation models demand efficient data analytic frameworks for scaling them to large geographic regions. Agent-based models are widely used to observe and analyze the urban growth simulation at various scales. The incorporation of the agent-based model makes the scaling task even harder due to communication and coordination among agents. Many existing agent-based model frameworks were implemented using traditional shared and distributed memory programming models. On the other hand, Apache Spark is becoming a popular platform for distributed big data in-memory analytics. This paper presents an implementation of agent-based sub-model in Apache Spark framework. With the in-memory computation, Spark implementation outperforms the traditional distributed memory implementation using MPI. This paper provides (i) an overview of our framework capable of running urban growth simulations at a fine resolution of 30 meter grid cells, (ii) a scalable approach using Apache Spark to implement an agent-based model for simulating human decisions, and (iii) the comparative analysis of performance of Apache Spark and MPI based implementations.","PeriodicalId":416086,"journal":{"name":"International Workshop on Analytics for Big Geospatial Data","volume":"2017 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-10-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127563550","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
G. Câmara, L. F. Assis, G. R. Queiroz, K. Ferreira, E. Llapa, L. Vinhas
Earth observation satellites produce petabytes of geospatial data. To manage large data sets, researchers need stable and efficient solutions that support their analytical tasks. Since the technology for big data handling is evolving rapidly, researchers find it hard to keep up with the new developments. To lower this burden, we argue that researchers should not have to convert their algorithms to specialised environments. Imposing a new API to researchers is counterproductive and slows down progress on big data analytics. This paper assesses the cost of research-friendliness, in a case where the researcher has developed an algorithm in the R language and wants to use the same code for big data analytics. We take an algorithm for remote sensing time series analysis on compare it use on map/reduce and on array database architectures. While the performance of the algorithm for big data sets is similar, organising image data for processing in Hadoop is more complicated and time-consuming than handling images in SciDB. Therefore, the combination of the array database SciDB and the R language offers an adequate support for researchers working on big Earth observation data analytics.
{"title":"Big earth observation data analytics: matching requirements to system architectures","authors":"G. Câmara, L. F. Assis, G. R. Queiroz, K. Ferreira, E. Llapa, L. Vinhas","doi":"10.1145/3006386.3006393","DOIUrl":"https://doi.org/10.1145/3006386.3006393","url":null,"abstract":"Earth observation satellites produce petabytes of geospatial data. To manage large data sets, researchers need stable and efficient solutions that support their analytical tasks. Since the technology for big data handling is evolving rapidly, researchers find it hard to keep up with the new developments. To lower this burden, we argue that researchers should not have to convert their algorithms to specialised environments. Imposing a new API to researchers is counterproductive and slows down progress on big data analytics. This paper assesses the cost of research-friendliness, in a case where the researcher has developed an algorithm in the R language and wants to use the same code for big data analytics. We take an algorithm for remote sensing time series analysis on compare it use on map/reduce and on array database architectures. While the performance of the algorithm for big data sets is similar, organising image data for processing in Hadoop is more complicated and time-consuming than handling images in SciDB. Therefore, the combination of the array database SciDB and the R language offers an adequate support for researchers working on big Earth observation data analytics.","PeriodicalId":416086,"journal":{"name":"International Workshop on Analytics for Big Geospatial Data","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-10-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128180423","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Given a region S comprised of locations that each have a time series of length |T|, the Persistent Change Windows (PCW) discovery problem aims to find all spatial window and temporal interval pairs that exhibit persistent change of attribute values over time. PCW discovery is important for critical societal applications such as detecting desertification, deforestation, and monitoring urban sprawl. The PCW discovery problem is challenging due to the large number of candidate patterns, the lack of monotonicity where sub-regions of a PCW may not show persistent change, the lack of predefined window sizes for the ST windows, and large datasets of detailed resolution and high volume, i.e., spatial big data. Previous approaches in ST change footprint discovery have focused on local spatial footprints for persistent change discovery and may not guarantee completeness. In contrast, we propose a space-time window enumeration and pruning (SWEP) approach that considers zonal spatial footprints when finding persistent change patterns. We provide theoretical analysis of SWEP's correctness, completeness, and space-time complexity. We also present a case study on vegetation data that demonstrates the usefulness of the proposed approach. Experimental evaluation on synthetic data show that the SWEP approach is orders of magnitude faster than the naive approach.
{"title":"Discovering persistent change windows in spatiotemporal datasets: a summary of results","authors":"Xun Zhou, S. Shekhar, Dev Oliver","doi":"10.1145/2534921.2534928","DOIUrl":"https://doi.org/10.1145/2534921.2534928","url":null,"abstract":"Given a region S comprised of locations that each have a time series of length |T|, the Persistent Change Windows (PCW) discovery problem aims to find all spatial window and temporal interval pairs <Si, Ti> that exhibit persistent change of attribute values over time. PCW discovery is important for critical societal applications such as detecting desertification, deforestation, and monitoring urban sprawl. The PCW discovery problem is challenging due to the large number of candidate patterns, the lack of monotonicity where sub-regions of a PCW may not show persistent change, the lack of predefined window sizes for the ST windows, and large datasets of detailed resolution and high volume, i.e., spatial big data. Previous approaches in ST change footprint discovery have focused on local spatial footprints for persistent change discovery and may not guarantee completeness. In contrast, we propose a space-time window enumeration and pruning (SWEP) approach that considers zonal spatial footprints when finding persistent change patterns. We provide theoretical analysis of SWEP's correctness, completeness, and space-time complexity. We also present a case study on vegetation data that demonstrates the usefulness of the proposed approach. Experimental evaluation on synthetic data show that the SWEP approach is orders of magnitude faster than the naive approach.","PeriodicalId":416086,"journal":{"name":"International Workshop on Analytics for Big Geospatial Data","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-11-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117186161","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Aggregate operators are a useful class of operators in relational databases. In this paper, we examine spatial aggregate operators over regions. Spatial aggregates are defined to operate over a set of regions, and return a single region as a result. We systematically identify individual spatial aggregate operations by extending existing spatial operations into aggregate form. Semantic meaning for each operator is defined over a specified data model. Once defined, algorithms for computing spatial aggregates over regions are provided. We show that all proposed aggregates can be computed using a single algorithm. Furthermore, we provide serial and parallel algorithm constructions that can take advantage of vector co-processors, such as graphical processing units (GPUs), and that can be integrated into map/reduce queries to take advantage of big data-style clusters. Example queries and their results are provided.
{"title":"Algorithms for fundamental spatial aggregate operations over regions","authors":"Mark McKenney, Brian Olsen","doi":"10.1145/2534921.2534930","DOIUrl":"https://doi.org/10.1145/2534921.2534930","url":null,"abstract":"Aggregate operators are a useful class of operators in relational databases. In this paper, we examine spatial aggregate operators over regions. Spatial aggregates are defined to operate over a set of regions, and return a single region as a result. We systematically identify individual spatial aggregate operations by extending existing spatial operations into aggregate form. Semantic meaning for each operator is defined over a specified data model. Once defined, algorithms for computing spatial aggregates over regions are provided. We show that all proposed aggregates can be computed using a single algorithm. Furthermore, we provide serial and parallel algorithm constructions that can take advantage of vector co-processors, such as graphical processing units (GPUs), and that can be integrated into map/reduce queries to take advantage of big data-style clusters. Example queries and their results are provided.","PeriodicalId":416086,"journal":{"name":"International Workshop on Analytics for Big Geospatial Data","volume":"559 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-11-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124174957","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：