Data driven insights are integral to new forms of knowledge generation in data rich settings. Specifically, in an era of new access to movement data, researchers have opportunities to leverage exis...

Cartographic map generalization involves complex rules, and a full automation has still not been achieved, despite many efforts over the past few decades. Pioneering studies show that some map generalization tasks can be partially automated by deep neural networks (DNNs). However, DNNs are still used as black-box models in previous studies. We argue that integrating explainable AI (XAI) into a DL-based map generalization process can give more insights to develop and refine the DNNs by understanding what cartographic knowledge exactly is learned. Following an XAI framework for an empirical case study, visual analytics and quantitative experiments were applied to explain the importance of input features regarding the prediction of a pre-trained ResU-Net model. This experimental case study finds that the XAI-based visualization results can easily be interpreted by human experts. With the proposed XAI workflow, we further find that the DNN pays more attention to the building boundaries than the interior parts of the buildings. We thus suggest that boundary intersection over union is a better evaluation metric than commonly used intersection over union in qualifying raster-based map generalization results. Overall, this study shows the necessity and feasibility of integrating XAI as part of future DL-based map generalization development frameworks.

Landmarks play key roles in human wayfinding and mobile navigation systems. Existing computational landmark selection models mainly focus on outdoor environments, and aim to identify suitable landmarks for guiding users who are unfamiliar with a particular environment, and fail to consider familiar users. This study proposes a familiarity-dependent computational method for selecting suitable landmarks for communicating with familiar and unfamiliar users in indoor environments. A series of salience measures are proposed to quantify the characteristics of each indoor landmark candidate, which are then combined in two LambdaMART-based learning-to-rank models for selecting landmarks for familiar and unfamiliar users, respectively. The evaluation with labelled landmark preference data by human participants shows that people’s familiarity with environments matters in the computational modelling of indoor landmark selection for guiding them. The proposed models outperform state-of-the-art models, and achieve hit rates of 0.737 and 0.786 for familiar and unfamiliar users, respectively. Furthermore, semantic relevance of a landmark candidate is the most important measure for the familiar model, while visual intensity is most informative for the unfamiliar model. This study enables the development of human-centered indoor navigation systems that provide familiarity-adaptive landmark-based navigation guidance.

Big GPS trajectory datasets can have redundant spatio-temporal information for applications, which requires simplification as a key preprocessing for modeling. Many existing simplification methods focus on the geometric information from a trajectory per se. Conversely, methods considering geographic context often fail to provide spatially adaptive simplification, or require complex parameter settings to achieve this task. This study proposes a novel two-stage adaptive trajectory simplification method embedding spatial indexing, enrichment, and aggregation in an integrated process. The first stage employs a quadtree for the subdivision depending on the density of geographic context features (i.e. POIs), leading to a variable-resolution representation of the area. The second stage aggregates trajectory waypoints locating in the same quadtree leaf node into a representative point, making the aggregation adapting to the spatial layout of the geographic feature in the first stage. Evaluation with a real-world vehicle trajectory dataset shows that the proposed approach can automatically simplify trajectory segments at variable compression ratios with greater simplification in areas with sparse context features (e.g. rural) and less simplification in areas with dense context features (e.g. urban). More importantly, the method can still preserve inter-trajectory distances between original trajectories and simplified ones, while significantly reducing the computing time.

Meng Huangab , Zhixiang Fangac*, Robert Weibelb , Tao Zhangd & Haosheng Huange* a State Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University, Wuhan, Hubei, PR Chinab Department of Geography, University of Zurich, Zurich, Switzerlandc Collaborative Innovation Center of Geospatial Technology, Wuhan, PR Chinad China Mobile Group Hubei Company Limited, Wuhan, Chinae Department of Geography, Ghent University, Ghent, BelgiumCONTACT Zhixiang Fang [email protected] Huang [email protected] billboards, as a new form of outdoor advertising, has gained popularity in recent years per its revolutionized way to control when and where the specific ads appear. However, this development also demands more complicated optimization for strategic deployments: the advertisers have to not only decide on a set of locations to display their ads, but also when to display them. The existing static optimization approaches become insufficient for this dynamic scenario to match advertisement and intended audience. Therefore, this research proposes three models in a workflow to mine mobile phone data and points of interest (POIs) data and to meet advertising needs in various situations. The three optimization models include a dynamic audience model to maximize the coverage of the target users, a dynamic environment model to maximize the coverage of the target environment, and a dynamic integrated model to maximize the coverage of both target audience and environment. A case study using shopping ads in Wuxue, China tests the three optimalization models. The results show that the proposed models are effective for providing an optimal solution for digital billboard configuration with a greater coverage of the target audience and environment compared to the state-of-the-art static models.

There has not been a time in the history of GIScience when movement analytics and mobility insights have played such an important role in policymaking as in today’s global responses to the COVID-19...

“We would like to dedicate this special issue to the memory of Professor Rein Ahas (1966 – 2018), a pioneer of computational movement analysis and mobility analytics, who sadly and very unexpectedl...

The study of movement is progressing rapidly as a subdiscipline in Geographic Information Science (GIScience). At the fulcrum of this new research area in GIScience are movement observations. Movem...

For applications in animal movement, we propose a random trajectory generator (RTG) algorithm that combines the concepts of random walks, space-time prisms, and the Brownian bridge movement model and is capable of efficiently generating random trajectories between a given origin and a destination point, with the least directional bias possible. Since we provide both a planar and a spherical version of the algorithm, it is suitable for simulating trajectories ranging from the local scale up to the (inter-)continental scale, as exemplified by the movement of migrating birds. The algorithm accounts for physical limitations, including maximum speed and maximum movement time, and provides the user with either single or multiple trajectories as a result. Single trajectories generated by the RTG algorithm can be used as a null model to test hypotheses about movement stimuli, while the multiple trajectories can be used to create a probability density surface akin to Brownian bridges.

This article describes a novel approach for finding similar trajectories, using trajectory segmentation based on movement parameters (MPs) such as speed, acceleration, or direction. First, a segmentation technique is applied to decompose trajectories into a set of segments with homogeneous characteristics with respect to a particular MP. Each segment is assigned to a movement parameter class (MPC), representing the behavior of the MP. Accordingly, the segmentation procedure transforms a trajectory to a sequence of class labels, that is, a symbolic representation. A modified version of edit distance called normalized weighted edit distance (NWED) is introduced as a similarity measure between different sequences. As an application, we demonstrate how the method can be employed to cluster trajectories. The performance of the approach is assessed in two case studies using real movement datasets from two different application domains, namely, North Atlantic Hurricane trajectories and GPS tracks of couriers in London. Three different experiments have been conducted that respond to different facets of the proposed techniques and that compare our NWED measure to a related method.

The introduction of automated generalisation procedures in map production systems requires that generalisation systems are capable of processing large amounts of map data in acceptable time and that cartographic quality is similar to traditional map products. With respect to these requirements, we examine two complementary approaches that should improve generalisation systems currently in use by national topographic mapping agencies. Our focus is particularly on self‐evaluating systems, taking as an example those systems that build on the multi‐agent paradigm. The first approach aims to improve the cartographic quality by utilising cartographic expert knowledge relating to spatial context. More specifically, we introduce expert rules for the selection of generalisation operations based on a classification of buildings into five urban structure types, including inner city, urban, suburban, rural, and industrial and commercial areas. The second approach aims to utilise machine learning techniques to extract heuristics that allow us to reduce the search space and hence the time in which a good cartographical solution is reached. Both approaches are tested individually and in combination for the generalisation of buildings from map scale 1:5000 to the target map scale of 1:25 000. Our experiments show improvements in terms of efficiency and effectiveness. We provide evidence that both approaches complement each other and that a combination of expert and machine learnt rules give better results than the individual approaches. Both approaches are sufficiently general to be applicable to other forms of self‐evaluating, constraint‐based systems than multi‐agent systems, and to other feature classes than buildings. Problems have been identified resulting from difficulties to formalise cartographic quality by means of constraints for the control of the generalisation process.

Web service technologies can be used to establish an interoperable framework between different generalisation systems. In a previous article three categories of generalisation web services were identified, including support services, operator services and processing services. This paper focuses on the category of support services. In a service‐based generalisation system, the purpose of support services is to assist the generalisation process by providing auxiliary measures, procedures and data structures that allow the representation of structural cartographic knowledge. The structural knowledge of the spatial and semantic context and the modelling of structural and spatial relationships is critical for the understanding of the role of cartographic features and thus for automated generalisation. Support services should extract and model this knowledge from the raw data and make it available to other generalisation operators. On the one hand the structural knowledge can be expressed by enriching map features with additional geometries or attributes. On the other hand, there exist various hierarchical and non‐hierarchical relationships between map features, many of which can be represented by graph data structures. After a brief introduction to the interoperable web service framework, this paper proposes a taxonomy of generalisation support services and discusses its elements. It is then shown how the complex output of such services can be represented for use with web services and stored in a reusable fashion. Finally, the utilisation of support services is illustrated on four implementation examples of support services that also highlight the interactions with the generalisation operators that use these auxiliary services.

This paper proposes and implements a new methodology for progressive transmission of vector lines and polygons over the Internet. The methodology generates continuous vector data through constrained remove operations of vertices on the server side, maintains consistent topology via a set of constraint rules, and restores original vector data through a reconstruction operator on the client side. The prototype system was implemented to investigate the performance of the methodology in terms of the preservation of consistent topology, transmission time, and qualities of the resulting visualizations. The method is shown to be scalable to large data sets, to produce graphically acceptable results and to maintain topology.

In this paper we describe work on the automatic recognition of island structures. In an initial phase several test persons were asked to mark groups of islands that they perceived on test maps. Based on these experimental results the island structures were categorized with respect to size and shape, and their construction described using principles from Gestalt theory. Based on those descriptions of island structures we will present an algorithm for the detection of large groups of islands based on a Minimal Spanning Tree (MST). Therefore, we apply split and merge operations on the MST. For the automated characterization of the shape and orientation of island groups we propose to use principal components obtained from a PCA. The results of the algorithm are then visually compared with the island groups previously marked by test persons and shortcomings of the approach are discussed.

Displacement, an operation of cartographic generalization, resolves congestion and overlap of map features that is caused by enlargement of map symbols to ensure readability at reduced scales. Algorithms for displacement must honour spatial context, avoid creating secondary spatial conflicts, and retain spatial patterns and relations such as alignments and relative distances that characterize the original map features. We present an algorithm for displacement of buildings based on optimization. While existing approaches directly displace the individual buildings, our algorithm first forms a truss of of elastic beams to capture important spatial patterns and preserve them during displacement. The algorithm proceeds in two phases. The first phase analyses spatial relationships to construct a truss as a weighted graph. The truss is initially based on the minimum spanning tree connecting the building centroids, with beam stiffness determined by spatial relationships. The second phase iteratively deforms the truss to minimize energy until a user‐defined distance is achieved. At each iteration, it computes forces on the truss, calculates truss deformations, and adjusts all build positions simultaneously. A prototype has been implemented to demonstrate the feasibility of the approach. The results are cartographically pleasing; in particular, spatial relationships between buildings are preserved.

Technological advances in position‐aware devices are leading to a wealth of data documenting motion. The integration of spatio‐temporal data‐mining techniques in GIScience is an important research field to overcome the limitations of static Geographic Information Systems with respect to the emerging volumes of data describing dynamics. This paper presents a generic geographic knowledge discovery approach for exploring the motion of moving point objects, the prime modelling construct to represent GPS tracked animals, people, or vehicles. The approach is based on the concept of geospatial lifelines and presents a formalism for describing different types of lifeline patterns that are generalizable for many application domains. Such lifeline patterns allow the identification and quantification of remarkable individual motion behaviour, events of distinct group motion behaviour, so as to relate the motion of individuals to groups. An application prototype featuring novel data‐mining algorithms has been implemented and tested with two case studies: tracked soccer players and data points representing political entities moving in an abstract ideological space. In both case studies, a set of non‐trivial and meaningful motion patterns could be identified, for instance highlighting the characteristic ‘offside trap’ behaviour in the first case and identifying trendsetting districts anticipating a political transformation in the latter case.

No abstract available.

Current geographical information systems (GIS) include functions to create cartographic output interactively or automatically. However, none of the systems presently incorporates mechanisms to ensure correct application of graphic functions. Many of today's users of GIS are not trained in the design and production of graphics and may create maps that confuse or mislead map users. Poorly-designed maps may obliterate the patterns in displayed information. This article investigates ways to improve the quality of GIS graphical displays and increase their effectiveness for decision-making. Alternatives are presented for improving map design in GIS, ranging from non-technical efforts to enhancement of software functionality. Suggestions propose augmenting users' abilities to create maps in an automated environment, using an approach termed amplified intelligence: key decisions default explicitly to users, whose knowledge is amplified by a range of high-level tools to carry out map design operations automatically.

This paper reviews the prospects of computer-assisted generalization of spatial data. Generalization as a general human activity is first considered in a broad context and map generalization is defined as a special variant of spatial modelling. It is then argued that in computer-assisted generalization, the spatial modelling process can be simulated only by strategies based on understanding and not by a mere sequence of operational processing steps. A conceptual framework for knowledge-based generalization is then presented which can be broken down into five steps: structure recognition, process recognition, process modelling, process execution and display. With reference to the goals of map generalization we identified tasks of statistical and cartographic generalization. The use of these types of tasks is discussed in relation to the concepts of digital landscape models (DLM) and digital cartographic models (DCM). A literature review is then presented in the context of this conceptual framework. It considers theoretical aspects of generalization and technical procedures on attributes and geometrical generalization. Specific sections of this review include statistical generalization, structure recognition and processes of cartographic generalization (point, line, area features, surfaces) and efforts for system integration. The paper concludes with an evaluation of the state of the art and an outlook on the future. Major efforts have to be devoted to developing an understanding of structures and processes involved in generalization (structure recognition, process recognition) and to modelling these processes. New data models will be a prerequisite for success in this field.