Spatial Knowledge Representation Research Articles

An agent architecture for expressive spatial knowledge and reasoning in land use modeling and simulations

This paper presents a new cognitive agent design approach integrating spatial knowledge representation and reasoning in agent-based modeling dedicated to land use simulations. A deep motivation for our agent-centric contribution is the ever-increasing development of spatially explicit agent simulation platforms. We build on this technological evolution and topology theory to endow the agent with human’s spatial representation and reasoning following a Belief–Desire–Intention architecture. A pilot implementation of the methodology with simulation experiments on a hunting model was carried out in GAMA platform to assess agent performances. Simulations display a consistent trend of animal population dynamics and also confirm a high model sensitivity to the integration of spatial knowledge and reasoning in agent-based models of human actor. These results demonstrate a successful implementation and the importance of spatial dimension in the expressive power and the validity of agent-based models. Future research efforts should therefore emphasize on designing human knowledge representation and incorporating learning abilities to improve models efficiency.

Assessing Spatial Knowledge and Mental Map Development Under Virtual Training Conditions

Mental maps are the core of wayfinding because they contain representations of point, route, and configuration related spatial knowledge. Currently, methods spatial knowledge do not clearly separate these theoretically disparate facets. The purpose of this research is to examine the relationships between point-related, route-related, and configuration-related spatial knowledge tests administered to participants in the context of training to comprehend a virtual environment. Results indicated that measures which are often interchanged under the assumption that they are reflective of spatial knowledge may in fact measure different and unrelated constructs. Futher, analyses showed that measures that are commonly used to tap point-related and route-related knowledge may have different influences on the development of configuration knowledge, and little to none on the ability for individuals to leverage their holistic mental map for wayfinding.

Fixing the inconsistencies of continuous changing operations in fuzzy spatiotemporal RDF graph

Fuzzy spatiotemporal data models have been used to support spatial and temporal knowledge representation and reasoning in the presence of fuzziness. In the meantime, Resource Description Framework (RDF), the most relevant data representation and exchange standard on Semantic Web, has become a trend for representing fuzzy spatiotemporal data. However, due to the dynamic changes of fuzzy spatiotemporal data, fuzzy spatiotemporal RDF graphs are more likely to violate predefined spatial and temporal constraints than the general RDF graph, which causes data inconsistencies. At present, consistency problems in RDF graphs have been widely studied, but their studies consider only discrete changing operations of fuzzy spatiotemporal data, while researches on consistencies of continuous changing operations are still open issues. In this paper, we study the types of inconsistency caused by continuous changing operations and propose the corresponding fixing rules to the problems of inconsistencies. Then, we propose the corresponding fixing approaches to checking and fixing fuzzy spatiotemporal RDF graphs according to the inconsistent states. Finally, the experimental results show that our proposed algorithms can significantly fix inconsistencies of continuous changing operations in fuzzy spatiotemporal RDF graphs.

Personalized Legibility of an Indoor Environment for People with Motor Disabilities: A New Framework

A mental map refers to the personalized representation of spatial knowledge in the human brain and is based on the perceptions, experiences, and interactions of people with their environment. For people with motor disabilities (PWMD) some perceptions and interactions with the environment during their mobility occur in different ways and consequently lead to different mental maps. For example, these people perceive and interact differently with elevators, escalators, and steps during their mobility. Hence, their perceptions of the level of complexity and the legibility of an environment may be different. Legibility of an environment is an indicator that measures the level of complexity and the ease of understanding of that environment by a person. In the literature, legibility is mostly estimated based on the environmental factors such as visibility, connectivity, and layout complexity for a given space. However, the role of personal factors (e.g., capacities) is rarely considered in the legibility assessment, which complicates its personalization. This paper aims at studying the influence of personal factors on the evaluation of the legibility of indoor environments for PWMD. In addition to the visibility, the connectivity, and the complexity of indoor environments, we also integrate the influence of the level of accessibility (i.e., presence of facilitators and obstacles) in the legibility assessment process. The Quebec City Convention Centre is selected as our study area and the legibility of this building is quantified. We show how the integration of the above-mentioned factors can influence the legibility for PWMD and hence their mobility performance in those environments.

Impact of Learning Methods on Spatial Knowledge Acquisition.

Research on the acquisition of spatial knowledge not only enriches our understanding of the theory of spatial knowledge representation but also creates practical value for the application of spatial knowledge. The aim of this study is to understand the impact of different learning methods on the acquisition of spatial knowledge, including the role of 2D maps, the difference between physical interaction and virtual interaction, and whether passive learning can replace active learning in virtual environments. One experiment was conducted, in which landmark knowledge and configurational knowledge were measured. Results indicate that 2D maps play a supporting role in acquiring both landmark knowledge and configurational knowledge. In addition, physical learning was associated with better spatial knowledge representation compared with virtual learning. An analysis of observational data in the third comparison found no significant difference between passive learning and active learning using virtual street view maps. However, with high-quality learning materials, passive learning can contribute to the acquisition of spatial knowledge more efficiently than active learning.

A Narrative Review of Theories of Wayfinding Within the Interior Environment.

The aim of this article is to provide a narrative overview of theories that are available in the published literature on wayfinding in interior environments. Wayfinding is an issue in healthcare facilities, and it has been documented that wayfinding problems have negative impacts on patients, visitors, caregivers' time, and the organization's bottom line. This issue can be more challenging for people with illness, low vision, cognitive disorders, and limited physical mobility. Wayfinding is a complex phenomenon depending on a variety of cognitive processes and behaviors. This article is a part of a broader literature review that searched for empirical studies on interior wayfinding. Four databases were systematically searched-PsychINFO, JSTOR, ProQuest, and EBSCO. A total of 81 articles satisfied all inclusion criteria for the broader study. Each article included in the broader study was reviewed to identify theory or theories, if any, that a study was founded on. After an in-depth review of the theories, the authors proposed four categories of theories relating to human wayfinding. The four classes and the specific theories in those are articulated in this article, including major changes during their evolution. The four classes of theories explain four facets of wayfinding: (1) theories of perception, (2) theories of spatial knowledge development, (3) theories of mental representation of spatial knowledge, and (4) theories of spatial cognition. Since wayfinding is essentially a cognitive/problem-solving phenomenon, understanding the theoretical underpinnings of wayfinding may result in more meaningful and impactful design decisions.

A fuzzy spatial description logic for the semantic web

Spatial information is a critical feature in a large number of application domains. Spatial information, however, is often not crisp but with the nature of imprecision and fuzziness. As the increasing requirements of spatial applications, there emerges many challenges regarding to the representation and reasoning of spatial knowledge. Description logic (DL) is a logical basis for representing knowledge and realizing reasoning tasks in the Semantic Web. Therefore, how to extend DL to achieve the goal of representing and reasoning fuzzy spatial knowledge needs to be settled. In this work, we study a fuzzy spatial extension of the well known fuzzy $$\mathcal {ALC}$$ DL to reason fuzzy spatial knowledge. First, we construct a fuzzy spatial concrete domain $$\mathcal {S}$$ which is comprised of fuzzy spatial regions and fuzzy RCC relationships. More importantly, we give the admissibility proof of fuzzy spatial concrete domain $$\mathcal {S}$$. Then we extend fuzzy $$\mathcal {ALC}$$ with an admissible fuzzy spatial concrete domain $$\mathcal {S}$$ and present a fuzzy spatial description logic f-$${\mathcal {ALC}}({\mathcal S})$$. Finally, we address a decision procedure for f-$${\mathcal {ALC}}({\mathcal S})$$ ABox consistency problem. Also, we show that the decision procedure is correct and the consistency problem for f-$${\mathcal {ALC}}({\mathcal S})$$ is decidable in PSPACE-complete.

Compact geometric representation of qualitative directional knowledge

To effectively and efficiently deal with large-scale spatial data is critical for applications in the age of information technology. Compact representation of spatial knowledge is one of the emerging research techniques that contribute to this capability. In this article, we consider the problem of compactly representing qualitative directional relations between extended objects, modelled in the Cardinal Direction Calculus (CDC) of Goyal and Egenhofer. For a large dataset of regions, this approach first constructs a simplified geometry for each region, which preserves CDC relations between regions, and then represents each simplified geometry compactly, so that the storage size is small while retrieving CDC relations from the representation is still reasonably fast. More specifically, the method called necessary cut is used to construct simple geometries, and the two methods, viz. the polygon representation and the rectangle representation, are devised to compactly represent the constructed geometries in cubic time w.r.t. the size of the corresponding simple geometry. Theoretical analyses demonstrate that the two representations, especially the rectangle representation, are promising to have small storage size. Moreover, our empirical evaluations on real-world datasets show that, for each dataset the new approach can produce a rectangle representation that has dominant performance against the state of the art techniques in reducing the storage size of the relations, while the average efficiency of retrieving CDC relations based on the rectangle representation is about the same as the fastest method in the literature.

Mobile Robot Navigation Based on Regionalized Spatial Knowledge Representation and Reasoning

Abstract A regionalized environmental knowledge model (REKModel) is presented to describe the environment in the paper. The REKModel is a hierarchical structure in which small regions are grouped together to form superordinate regions. The REKModel is intrinsically hierarchical iterative and nested. Thus an extended nested-graph(ENG) is proposed to construct REKModel. An biomimetic navigation system for mobile robots is presented that is inspired by the fine-to-coarse planning heuristic, a human wayfinding strategy. A online fine-to-coarse pathfinding algorithm designed here allows robots to derives the route with decreasing the level of detail along the route. By using spatial information at different levels of detail for close and coarseness for distance, the memory load and plan complexity are all reduced. The simulation on MobileSim platform verifies the effectiveness and feasibility of the navigation system.

Understanding landmark types in human navigation from a database perspective

Abstract. Delivering location based information to support human navigation in unfamiliar environments has been a long standing research topic in Location-based Service (LBS). Enriching route and direction communication with landmark information not only increases the efficiency of navigation applications but also facilitates spatial knowledge learning among users (Lovelace et al. 1999, Wunderlich & Gramann 2018). Landmarks are salient geographic entities anchored at fixed positions in an environment. In wayfinding and navigation, landmarks serve for multiple purposes as and organizing concept that help people build up structured spatial knowledge representation and as navigational tools to support navigational decision-making (Sorrows & Hirtle 1999). In previous work, landmarks are often conceptualized as reference points. However, in Geodatabase landmarks are not always represented as point features. Moreover, non-point features are also important references in route instructions, especially for pedestrian navigation. The signs, though often considered as point elements not only provide something about “here”, but also directional information (Gibson 2009). In this work, we attempt to provide a categorization of landmark for route communications from a database perspective. The proposed view is intended to be extended to a landmark model to generate various route communication, such as verbal instructions or symbolic visualizations.

2D-to-3D Scene Generation and Rendering using CNN and Spatial Knowledge Representation

2D-to-3D Scene Generation and Rendering using CNN and Spatial Knowledge Representation

Acquiring Common Sense Spatial Knowledge Through Implicit Spatial Templates

Spatial understanding is a fundamental problem with wide-reaching real-world applications. The representation of spatial knowledge is often modeled with spatial templates, i.e., regions of acceptability of two objects under an explicit spatial relationship (e.g., "on," "below," etc.). In contrast with prior work that restricts spatial templates to explicit spatial prepositions (e.g., "glass on table"), here we extend this concept to implicit spatial language, i.e., those relationships (generally actions) for which the spatial arrangement of the objects is only implicitly implied (e.g., "man riding horse"). In contrast with explicit relationships, predicting spatial arrangements from implicit spatial language requires significant common sense spatial understanding. Here, we introduce the task of predicting spatial templates for two objects under a relationship, which can be seen as a spatial question-answering task with a (2D) continuous output ("where is the man w.r.t. a horse when the man is walking the horse?"). We present two simple neural-based models that leverage annotated images and structured text to learn this task. The good performance of these models reveals that spatial locations are to a large extent predictable from implicit spatial language. Crucially, the models attain similar performance in a challenging generalized setting, where the object-relation-object combinations (e.g., "man walking dog") have never been seen before. Next, we go one step further by presenting the models with unseen objects (e.g., "dog"). In this scenario, we show that leveraging word embeddings enables the models to output accurate spatial predictions, proving that the models acquire solid common sense spatial knowledge allowing for such generalization.

A representation of partial spatial knowledge: a cognitive map approach for evacuation simulations

ABSTRACTUsually, routing models in evacuation simulations assume that agents have comprehensive and global knowledge about the building's structure. They neglect the fact that pedestrians might possess no or only parts of information about their position relative to final exits and possible routes leading to them. For the sake of a more realistic description of the routing process, we introduce the systematics of using partial spatial knowledge. Particularly, we present an agent-based approach modeling the inaccurate mental representation of pedestrians' spatial knowledge (the cognitive map). In addition, the model considers further principles and constraints of human wayfinding. Furthermore, we present results of a field study which we conducted in an office building. The purpose of this study was to investigate route choices of people in dependency on their familiarity with the building. Our modeling approach is then calibrated using the obtained results. In this context, the distribution of routes which were used by the subjects are compared with results of the model.

Indoors Locality Positioning Using Cognitive Distances and Directions.

Spatial relationships are crucial to spatial knowledge representation, such as positioning localities. However, minimal attention has been devoted to positioning localities indoors with locality description. Distance and direction relations are generally used when positioning localities, namely, translating descriptive localities into spatially explicit ones. We propose a joint probability function to model locality distribution to address the uncertainty of positioning localities. The joint probability function consists of distance and relative direction membership functions. We propose definitions and restrictions for the use of the joint probability function to make the locality distribution highly practical. We also evaluate the performance of our approach through indoor experiments. Test results demonstrate that a positioning accuracy of 3.5 m can be achieved with the semantically derived spatial relationships.

Questions of uncertainty in geography

The concept of uncertainty has fostered in the last decade’s fundamental and applied research in different disciplinary fields. Couclelis (2003) clearly demonstrated the pervasiveness of uncertainty in the production process of geographical knowledge. The paper shares this epistemological point of view. Pragmatically, its goal is to show how questions of uncertainty arise in the praxis of geographic research. It suggests that scientific work can be enriched, and not hindered, by addressing uncertainty in knowledge. The paper discusses eight domains within the activity of the geographer, where questions of uncertainty arise: geographic information, geographic definitions, the explanation of geographic phenomena, the complexity of spatial systems, geosimulation, the representation of spatial knowledge, subjectivity in spatial phenomena, and planning. Within each domain uncertainty issues are identified as well as their possible interrelations.

Salience Indicators for Landmark Extraction at Large Spatial Scales Based on Spatial Analysis Methods

Urban landmarks are frequently used in way-finding and representations of spatial knowledge. However, assessing the salience of urban landmarks is difficult. Moreover, no method exists to rapidly extract urban landmarks from basic geographic information databases. The goal of this paper is to solve these problems from the dual aspects of spatial knowledge representation and public spatial cognition rules. A clear and systematic definition for multiple-scale urban landmarks is proposed, together with a category reference for extracting small- and medium-scale urban landmarks and a model for the large-scale automatic extraction of urban landmarks. In this large-scale automatic urban landmark extraction model, the salience is expressed by two weighted parameters: the check-in totals and local accessibility. The extraction threshold is set according to a predefined number of landmarks to be extracted. Experiments show that the extraction results match the reference data well.

Fixing inconsistencies of fuzzy spatiotemporal XML data

Fuzzy spatiotemporal data models have been used to support spatial and temporal knowledge representation and reasoning in the presence of fuzziness. In the meantime, XML is expected to become the next generation standard language for exchanging data over the Internet, which will become a trend to represent fuzzy spatiotemporal data based on XML. However, fuzzy spatiotemporal XML documents may contain inconsistencies violating predefined spatial and temporal constraints, which cause the data inconsistency problems. Although those consistency problems in XML documents have been widely studied, their studies only take the general data into account, and the studies on consistencies of fuzzy spatiotemporal data are still open issues. In this paper we put forward solutions to the problems of inconsistencies in fuzzy spatiotemporal XML documents. We also analyze inconsistent states which are named discontinuity overlap or cycle of the temporal labels of some incoming edges. Then, we put forward the corresponding approaches to checking and fixing fuzzy spatiotemporal XML documents according to the inconsistent states. Finally, the experimental results show that our proposed algorithms can fix inconsistencies of fuzzy spatiotemporal XML documents significantly.

Constructing representations of spatial location from briefly presented displays.

Spatial memory and reasoning rely heavily on allocentric (often map-like) representations of spatial knowledge. While research has documented many ways in which spatial information can be represented in allocentric form, less is known about how such representations are constructed. For example: Are the very early, pre-attentive parts of the process hard-wired, or can they be altered by experience? We addressed this issue by presenting sub-saccadic (53ms) masked stimuli consisting of a target among one to three reference features. We then shifted the location of the feature array, and asked participants to identify the target's new relative location. Experience altered feature processing even when the display duration was too short to allow attention re-allocation. The results demonstrate the importance of early perceptual processes in the creation of representations of spatial location, and the malleability of those processes based on experience and expectations.

Алгоритмізація геоінформаційного картографування міжнародних зв'язків

The objective of the article is to show the algorithm of international relations geoinformation mapping. During the mapping we can mine unknown or hidden knowledge which can be understood from the spatial database and improve the ability of interpreting data. The algorithm of mapping can be divided into five phases described as follows: investigation of potential map user’s demands, spatial data analysis and selection (accuracy, consistency, fullness); scale, map projections and map composition choice (main meridian, map frame, map distortions), spatial data preprocessing (metadata description, data normalization, data generalization, data quality control, error detection, anomalies, duplication and noise detection, correction of errors, data deduplication), data mining (creation of classifications, cluster and factor analysis, decision trees, neural networks, self-organizing maps, association rules, quality assessment), map creation, knowledge representation and evaluation. The article arranges experience of international relations geoinformation mapping. The international relations maps of Ukraine potential users (state and region management, educational, cultural, advocacy, international marketing and advertising, media business), subjects (substance, energy and information flows) and indicators arrays are proposed. Main subjects of maps can be described as follows: military-political, economicpolitical, socio-political and ideological relations; migrations and refugees, international tourism, export and import of goods and services, information flows (flows of culture elements, artifacts and tangible cultural monuments, scientific knowledge, ideas, technologies, traditions, beliefs, ideology, etc.), transport and business infrastructure. The requirements for spatial data sources for international relations geoinformation mapping are defined. The algorithm of international relations static and interactive maps creation and analysis are founded. The flow chart of Ukraine international relations geoinformation mapping algorithm is presented.

Computing force field-based directional maps in subquadratic time

Spatial knowledge, i.e., knowledge about configurations among distinct spatial entities, plays a key role in everyday life and has wide applications in numerous application domains. In particular, computer vision systems exploit spatial knowledge to enhance their object recognition capabilities. The focus of this article is on the representation of spatial knowledge. Specifically, we concentrate on the representation of spatial relationships among objects in an image, a fundamental problem in pattern recognition and computer vision. In the study of spatial relationships, significant progress has been made by using directional maps. However, the time complexity of map generation is O(N2) and thus not practical. We present an improvement of a state-of-the-art approximation method that is based on an analogy with the force exerted by an object on another object. The improvement is based on the use of a family of forces whose dependence on the pixel–pixel distance r is of the type 1/rl+1 with l > 1. Analysis of time complexity shows that the improved method is O(Nx), with x=1+3/(l+2). Application of this method to both artificial and real images validates the time complexity analysis and shows that the proposed method is much faster than the original method. Furthermore, for the special case of small objects relative to the image size, the proposed method generates more accurate maps and, in other cases, the difference in accuracy is very small.