SE‐KGE: A location‐aware Knowledge Graph Embedding model for Geographic Question Answering and Spatial Semantic Lifting

Learning knowledge graph (KG) embeddings is an emerging technique for a variety of downstream tasks such as summarization, link prediction, information retrieval, and question answering. However, most existing KG embedding models neglect space and, therefore, do not perform well when applied to (geo)spatial data and tasks. Most models that do consider space primarily rely on some notions of distance. These models suffer from higher computational complexity during training while still losing information beyond the relative distance between entities. In this work, we propose a location‐aware KG embedding model called SE‐KGE. It directly encodes spatial information such as point coordinates or bounding boxes of geographic entities into the KG embedding space. The resulting model is capable of handling different types of spatial reasoning. We also construct a geographic knowledge graph as well as a set of geographic query–answer pairs called DBGeo to evaluate the performance of SE‐KGE in comparison to multiple baselines. Evaluation results show that SE‐KGE outperforms these baselines on the DBGeo data set for the geographic logic query answering task. This demonstrates the effectiveness of our spatially‐explicit model and the importance of considering the scale of different geographic entities. Finally, we introduce a novel downstream task called spatial semantic lifting which links an arbitrary location in the study area to entities in the KG via some relations. Evaluation on DBGeo shows that our model outperforms the baseline by a substantial margin.     

GeoGraphViz: Geographically constrained 3D force-directed graph for knowledge graph visualization

Knowledge graphs are a key technique for linking and integrating cross-domain data, concepts, tools, and knowledge to enable data-driven analytics. As much of the world's data have become massive in size, visualizing graph entities and their interrelationships intuitively and interactively has become a crucial task for ingesting and better utilizing graph content to support semantic reasoning, discovering hidden knowledge discovering, and better scientific understanding of geophysical and social phenomena. Despite the fact that many such phenomena (e.g., disasters) have clear spatial footprints and geographic properties, their location information is considered only as a textual label in existing graph visualization tools, limiting their capability to reveal the geospatial distribution patterns of the graph nodes. In addition, most graph visualization techniques rely on 2D graph visualization, which constrains the dimensions of information that can be presented and lacks support for graph structure examination from multiple angles. To tackle the above challenges, we developed a novel 3D map-based graph visualization algorithm to enable interactive exploration of graph content and patterns in a spatially explicit manner. The algorithm extends a 3D force directed graph by integrating a web map, an additional geolocational force, and a force balancing variable that allows for the dynamic adjustment of the 3D graph structure and layout. This mechanism helps create a balanced graph view between the semantic forces among the graph nodes and the attractive force from a geolocation to a graph node. Our solution offers a new perspective in visualizing and understanding spatial entities and events in a knowledge graph.     

A Spatial Anomaly Points and Regions Detection Method Using Multi‐Constrained Graphs and Local Density

Spatial anomalies may be single points or small regions whose non‐spatial attribute values are significantly inconsistent with those of their spatial neighborhoods. In this article, a S patial A nomaly P oints and R egions D etection method using multi‐constrained graphs and local density ( SAPRD for short) is proposed. The SAPRD algorithm first models spatial proximity relationships between spatial entities by constructing a Delaunay triangulation, the edges of which provide certain statistical characteristics. By considering the difference in non‐spatial attributes of adjacent spatial entities, two levels of non‐spatial attribute distance constraints are imposed to improve the proximity graph. This produces a series of sub‐graphs, and those with very few entities are identified as candidate spatial anomalies. Moreover, the spatial anomaly degree of each entity is calculated based on the local density. A spatial interpolation surface of the spatial anomaly degree is generated using the inverse distance weight, and this is utilized to reveal potential spatial anomalies and reflect their whole areal distribution. Experiments on both simulated and real‐life spatial databases demonstrate the effectiveness and practicability of the SAPRD algorithm.     

A Spatio‐Temporal Graph Model for Marine Dune Dynamics Analysis and Representation

Defining a model for the representation and the analysis of spatio‐temporal dynamics remains an open domain in geographical information sciences. In this article we investigate a spatio‐temporal graph‐based model dedicated to managing and extracting sets of geographical entities related in space and time. The approach is based on spatial and temporal local relations between neighboring entities during consecutive times. The model allows us to extract sets of connected entities distant in time and space over long periods and large spaces. From GIS concepts and qualitative reasoning on space and time, we combine the graph model with a dedicated spatial database. It includes information on geometry and geomorphometric parameters, and on spatial and temporal relations. This allows us to extend classical measurements of spatial parameters, with comparisons of entities linked by complex relations in space and time. As a case study, we show how the model suggests an efficient representation of dunes dynamics on a nautical chart for safe navigation.     

Deeply integrating Linked Data with Geographic Information Systems

The realization that knowledge often forms a densely interconnected graph has fueled the development of graph databases, Web‐scale knowledge graphs and query languages for them, novel visualization and query paradigms, as well as new machine learning methods tailored to graphs as data structures. One such example is the densely connected and global Linked Data cloud that contains billions of statements about numerous domains, including life science and geography. While Linked Data has found its way into everyday applications such as search engines and question answering systems, there is a growing disconnect between the classical ways in which Geographic Information Systems (GIS) are still used today and the open‐ended, exploratory approaches used to retrieve and consume data from knowledge graphs such as Linked Data. In this work, we conceptualize and prototypically implement a Linked Data connector framework as a set of toolboxes for Esri's ArcGIS to close this gap and enable the retrieval, integration, and analysis of Linked Data from within GIS. We discuss how to connect to Linked Data endpoints, how to use ontologies to probe data and derive appropriate GIS representations on the fly, how to make use of reasoning, how to derive data that are ready for spatial analysis out of RDF triples, and, most importantly, how to utilize the link structure of Linked Data to enable analysis. The proposed Linked Data connector framework can also be regarded as the first step toward a guided geographic question answering system over geographic knowledge graphs.     

Knowledge and Reasoning in Spatial Analysis

Reasoning is an essential part of any analysis process. Especially in visual analytics, the quality of the results depends heavily on the knowledge and reasoning skills of the analyst. In this study, we consider how to make the results transparent by visualizing the reasoning and the knowledge, so that persons from outside can trace and verify them. The focus of this study is in spatial analysis and a case study was carried out on a process of off‐road mobility analysis. In the case study, linked views of a map and a PCP were identified as reasoning artifacts. The knowledge used by the analyst was formed by these artifacts and the tangible pieces of information identified in them, along with the mental models of the analyst′s mind. To make the results transparent, the tangible pieces of information were marked with sketches and the mental models were presented in causal graphs because it was found that causality was central to the reasoning process in the case study. The causal graph allows the reasoning of the analyst to be studied, as well as traced back to its origin.     

Fields as a framework for integrating GIS and environmental process models. Part 1: Representing spatial continuity

Linking a GIS to a spatially distributed, physically-based environmental model offers many advantages. However, the implementation of such linkages is generally problematic. Many problems arise because the relationship between the reality being represented by the mathematical model and the data model used to organize the spatial data in the GIS has not been rigorously defined. In particular, while many environmental models are based on theories that assume continuity and incorporate physical fields as independent variables, current GISs can only represent continuous phenomena in a variety of discrete data models. This paper outlines a strategy in which field variables are used to enable modellers to work directly with the spatial data as spatially continuous phenomena. This allows the manner in which the spatial data has been discretized and the ways in which it can be manipulated to be treated independently from the conceptual modelling of physical processes. Modellers can express their spatial data needs as representations of reality, rather than as elements of a GIS database, and a GIS-independent language for model development may result. By providing a formal linkage between the various models of spatial phenomena, a mechanism is created for the explicit expression of transformation rules between the different spatial data models stored and manipulated by a GIS.     

Simulating Complex Adaptive Geographic Systems: A Geographically Aware Intelligent Agent Approach

The objective of this paper is to present a spatially explicit agent-based simulation framework with a supporting software package to explore complex adaptive geographic systems. This framework is particularly suitable for modeling entities that are contextually aware, knowledge driven, and adaptive because it represents them as geographically aware intelligent agents. Fundamental advances in the explicit representation of contextual information, knowledge structures, and learning processes are needed for modeling intelligent agents situated within geographic systems. The representation of these agents requires the integration of agent-based models, machine learning, and GIS. Existing software packages for agent-based modeling, however, often provide insufficient support for this integration. The agent-based simulation package presented here is specifically designed to achieve such integration by assisting the development of agent-based models from the simulation framework. Object-oriented modeling techniques were used to implement this simulation package, which includes four modules: simulation, visualization, learning, and geoprocessing. In particular, the learning and geoprocessing modules facilitate the representation of adaptive behavior in agents within spatially explicit environments. The utility of the agent-based simulation package is illustrated using two simulation models: one of adaptive elk behavior and another of pedestrian movement. The successful design of the simulation models suggests that the modeling framework with the supporting software package is well suited to the resolution of complex adaptive geographic problems.     

Using GIS in Areawide Pest Management: A Case Study in South Dakota

Infestations of corn rootworms (Coleoptera: Chrysomelidae) create economic and environmental concerns in the Corn Belt region of the United States. To supplement the population control tactics of areawide pest management programs, we believe that a better understanding of the spatial relationships between biotic and abiotic or physical factors at the landscape scale is needed. Our research used several geographical information systems (GIS) and spatial analytical techniques to examine relationships between corn rootworm metapopulation dynamics, soil texture, and elevation. Within GIS, several spatially explicit procedures were used that include an interpolation technique, spatial autocorrelation analysis, and contingency analysis. Corn rootworm metapopulation distributions were found to be aggregated and related to soil texture and elevation. We review techniques and discuss our preferences for using particular spatially explicit procedures. The information derived from the spatial analyses demonstrates how GIS can be used in areawide pest management to provide inputs for spatially explicit models to predict future pest populations and formulate more well‐informed pest management decisions. The techniques described in this paper could easily be extended to study the spatial dynamics between other pest populations in agricultural landscapes.     

A Spatial Analysis System for Integrating Data,Methods and Models on Environmental Risks and Health Outcomes

Integrating data on health outcomes with methods of disease mapping and spatially explicit models of environmental contaminants are important aspects of environmental health surveillance. In this article, we describe a modular, web‐based spatial analysis system that uses GIS, spatial analysis methods and software services delivered over computer networks to achieve this end. The Environmental Health Surveillance System (EHSS) is a prototype system that is designed to serve three purposes: a secure environment for producing maps of disease outcomes from individual‐level data while preserving privacy; an automated process of linking environmental data, environmental models, and GIS tasks like geocoding for the purposes of estimating individual exposures to environmental contaminants; and mechanisms to visualize the spatial patterns of disease outcomes via Web‐based mapping interfaces and interactive tools like Google Earth.     

Spatially explicit models for exploring COVID‐19 lockdown strategies

This article describes two spatially explicit models created to allow experimentation with different societal responses to the COVID‐19 pandemic. We outline the work to date on modeling spatially explicit infective diseases and show that there are gaps that remain important to fill. We demonstrate how geographical regions, rather than a single, national approach, are likely to lead to better outcomes for the population. We provide a full account of how our models function, and how they can be used to explore many different aspects of contagion, including: experimenting with different lockdown measures, with connectivity between places, with the tracing of disease clusters, and the use of improved contact tracing and isolation. We provide comprehensive results showing the use of these models in given scenarios, and conclude that explicitly regionalized models for mitigation provide significant advantages over a “one‐size‐fits‐all” approach. We have made our models, and their data, publicly available for others to use in their own locales, with the hope of providing the tools needed for geographers to have a voice during this difficult time.     

Spatial Approaches to Modeling Dispersion of Communicable Diseases – A Review

The dispersion of communicable diseases in a population is intrinsically spatial. In the last several decades, a range of spatial approaches has been devised to model epidemiological processes; and they differ significantly from each other. A review of spatially oriented epidemiological models is necessary to assess advances in spatial approaches to modeling disease dispersion and to help identify those most appropriate for specific research goals. The most notable difference in the design of these spatially oriented models is the scale and mobility of the modeling unit. Using two criteria, this review identifies six types of spatially oriented models. These include: (1) population‐based wave models, (2) sub‐population models, (3) individual‐based cellular automata models, (4) mobile sub‐population models, (5) individual‐based spatially implicit models, and (6) individual‐based mobile models. Each model type is evaluated in terms of its design principles, assumptions, and intended applications. For the evaluation of design, four aspects of design principles are discussed: the modeling unit, the interaction between the modeling units, the spatial process, and the temporal process utilized in a design. Insights gained from this review can be useful for devising much‐needed spatially and temporally oriented strategies to forecast, prevent, and control communicable diseases.     

A Multiscale Approach for Spatio-Temporal Outlier Detection

A spatial outlier is a spatially referenced object whose thematic attribute values are significantly different from those of other spatially referenced objects in its spatial neighborhood. It represents an object that is significantly different from its neighbourhoods even though it may not be significantly different from the entire population. Here we extend this concept to the spatio‐temporal domain and define a spatial‐temporal outlier (ST‐outlier) to be a spatial‐temporal object whose thematic attribute values are significantly different from those of other spatially and temporally referenced objects in its spatial or/and temporal neighbourhoods. Identification of ST‐outliers can lead to the discovery of unexpected, interesting, and implicit knowledge, such as local instability or deformation. Many methods have been recently proposed to detect spatial outliers, but how to detect the temporal outliers or spatial‐temporal outliers has been seldom discussed. In this paper we propose a multiscale approach to detect ST‐outliers by evaluating the change between consecutive spatial and temporal scales. A four‐step procedure consisting of classification, aggregation, comparison and verification is put forward to address the semantic and dynamic properties of geographic phenomena for ST‐outlier detection. The effectiveness of the approach is illustrated by a practical coastal geomorphic study.     

Analytical queries on semantic trajectories using graph databases

This article studies the analysis of moving object data collected by location‐aware devices, such as GPS, using graph databases. Such raw trajectories can be transformed into so‐called semantic trajectories, which are sequences of stops that occur at “places of interest.” Trajectory data analysis can be enriched if spatial and non‐spatial contextual data associated with the moving objects are taken into account, and aggregation of trajectory data can reveal hidden patterns within such data. When trajectory data are stored in relational databases, there is an “impedance mismatch” between the representation and storage models. Graphs in which the nodes and edges are annotated with properties are gaining increasing interest to model a variety of networks. Therefore, this article proposes the use of graph databases (Neo4j in this case) to represent and store trajectory data, which can thus be analyzed at different aggregation levels using graph query languages (Cypher, for Neo4j). Through a real‐world public data case study, the article shows that trajectory queries are expressed more naturally on the graph‐based representation than over the relational alternative, and perform better in many typical cases.     

Computing and querying strict,approximate, and metrically refined topological relations in linked geographic data

Geographic entities and the information associated with them play a major role in Web‐scale knowledge graphs such as Linked Data. Interestingly, almost all major datasets represent places and even entire regions as point coordinates. There are two key reasons for this. First, complex geometries are difficult to store and query using the current Linked Data technology stack to a degree where many queries take minutes to return or will simply time out. Second, the absence of complex geometries confirms a common suspicion among GIScientists, namely that for many everyday queries place‐based relational knowledge is more relevant than raw geometries alone. To give an illustrative example, the statement that the White House is in Washington, DC is more important for gaining an understating of the city than the exact geometries of both entities. This does not imply that complex geometries are unimportant but that (topological) relations should also be extracted from them. As Egenhofer and Mark (1995b) put it in their landmark paper on naive geography, topology matters, metric refines. In this work we demonstrate how to compute and utilize strict, approximate, and metrically refined topological relations between several geographic feature types in DBpedia and compare our results to approaches that compute result sets for topological queries on the fly.     

Spatial‐Scene Similarity Queries

Assessing spatial scenes for similarity is difficult from a cognitive and computational perspective. Solutions to spatial‐scene similarity assessments are sensible only if corresponding elements in the compared scenes are identified correctly. This matching process becomes increasingly complex and error‐prone for large spatial scenes as it is questionable how to choose one set of associations over another or how to account quantitatively for unmatched elements. We develop a comprehensive methodology for similarity queries over spatial scenes that incorporates cognitively motivated approaches about scene comparisons, together with explicit domain knowledge about spatial objects and their relations for the relaxation of spatial query constraints. Along with a sound graph‐theoretical methodology, this approach provides the foundation for plausible reasoning about spatial‐scene similarity queries.     

Spatially constrained regionalization with multilayer perceptron

In this article, multilayer perceptron (MLP) network models with spatial constraints are proposed for regionalization of geostatistical point data based on multivariate homogeneity measures. The study focuses on non‐stationarity and autocorrelation in spatial data. Supervised MLP machine learning algorithms with spatial constraints have been implemented and tested on a point dataset. MLP spatially weighted classification models and an MLP contiguity‐constrained classification model are developed to conduct spatially constrained regionalization. The proposed methods have been tested with an attribute‐rich point dataset of geological surveys in Ukraine. The experiments show that consideration of the spatial effects, such as the use of spatial attributes and their respective whitening, improve the output of regionalization. It is also shown that spatial sorting used to preserve spatial contiguity leads to improved regionalization performance.     

Simulating Spatial Dynamics and Processes in a Retail Gasoline Market: An Agent‐Based Modeling Approach

Simulating the dynamics and processes within a spatially influenced retail market, such as the retail gasoline market, is a highly challenging research area. Current approaches are limited through their inability to model the impact of supplier or consumer behavior over both time and space. Agent‐based models (ABMs) provide an alternative approach that overcomes these problems. We demonstrate how knowledge of retail pricing is extended by using a ‘hybrid’ model approach: an agent model for retailers and a spatial interaction model for consumers. This allows the issue of spatial competition between individual retailers to be examined in a way only accessible to agent‐based models, allowing each model retailer autonomous control over optimizing their price. The hybrid model is shown to be successful at recreating spatial pricing dynamics at a national scale, simulating the effects of a rise in crude oil prices as well as accurately predicting which retailers were most susceptible to closure over a 10‐year period.     

A New GIS-based Solar Radiation Model and Its Application to Photovoltaic Assessments

The solar radiation model r.sun is a flexible and efficient tool for the estimation of solar radiation for clear‐sky and overcast atmospheric conditions. In contrast to other models, r.sun considers all relevant input parameters as spatially distributed entities to enable computations for large areas with complex terrain. Conceptually the model is based on equations published in the European Solar Radiation Atlas (ESRA). The r.sun model was applied to estimate the solar potential for photovoltaic systems in Central and Eastern Europe. The overcast radiation was computed from clear‐sky values and a clear‐sky index. The raster map of the clear‐sky index was computed using a multivariate interpolation method to account for terrain effects, with interpolation parameters optimized using a cross‐validation technique. The incorporation of terrain data improved the radiation estimates in terms of the model's predictive error and the spatial pattern of the model outputs. Comparing the results of r.sun with the ESRA database demonstrates that integration of the solar radiation model and the spatial interpolation tools in a GIS can be especially helpful for data at higher resolutions and in regions with a lack of ground measurements.     

Reconstructing facade semantic models using hierarchical topological graphs

Semantic information in 3D building models is of vital importance for various applications in terms of smart cities. To infer the semantic information and localize the components on building facades, this article proposes a novel approach to model facades with semantics by constructing hierarchical topological graphs. This method utilizes the topological characteristics of building facades. In the first‐layer layout graph, the algorithm takes the nearest cluster as the vertex and the distance between components as the edge. Thus, a topology graph is generated for the facade. The proposed algorithm is divided into three steps. First, the topology graph is obtained by calculating the spacing between the components. It is reasonable to calculate the topological graph by encoding the topological edges. If this calculation is not effective, the topology is justified by adjusting the spacing between components. Finally, the vertices in the graph are used to repair the occluded parts of the facade. In the second‐layer graph, a grid is constructed according to the first‐layer graph. Then, the attributes of the nodes are used to reconstruct the facade. The experimental results show that this method has a high accuracy of 90% and that the average time consumption is 6 s.