Catálogo de publicaciones - libros

Why are notions like semantics and ontologies suddenly getting so much attention, within and outside geospatial information communities? The main reason lies in the componentization of Geographic Information Systems (GIS) into services, which are supposed to interoperate within and across these communities. Consequently, I look at geospatial semantics in the context of semantic interoperability. The paper clarifies the relevant notion of semantics and shows what parts of geospatial information need to receive semantic specifications in order to achieve interoperability. No attempt at a survey of approaches to provide semantics is made, but a framework for solving interoperability problems is proposed in the form of semantic reference systems. Particular emphasis is put on the need and possible ways to ground geospatial semantics in physical processes and measurements.

Analysis of global geographic phenomena requires non-planar models. In the past, models for topological relations have focused either on a two-dimensional or a three-dimensional space. When applied to the surface of a sphere, however, neither of the two models suffices. For the two-dimensional planar case, the eight binary topological relations between spatial regions are well known from the 9-intersection model. This paper systematically develops the binary topological relations that can be realized on the surface of a sphere. Between two regions on the sphere there are three binary relations that cannot be realized in the plane. These relations complete the conceptual neighborhood graph of the eight planar topological relations in a regular fashion, providing evidence for a regularity of the underlying mathematical model. The analysis of the algebraic compositions of spherical topological relations indicates that spherical topological reasoning often provides fewer ambiguities than planar topological reasoning. Finally, a comparison with the relations that can be realized for one-dimensional, ordered cycles draws parallels to the spherical topological relations.

The main problem of visual query languages for geographical data concerns the query’s ambiguity. Ambiguity derives from the fact that a query can lead to multiple interpretations for both the system and user. In fact a query can have different visual representations, and these can themselves have different interpretations. Among the reasons leading to these ambiguities, one appears to be fundamental: the user gives his own semantics to the information. However his actions may not completely represent his intentions, so the system may make an incorrect interpretation. Additionally, when a user draws two icons representing different geographical objects of a query he cannot avoid defining one or more spatial relationships between them. This is the case for any pair of icons, however the user often does not want to define spatial relationships between all pair of icons. So he cannot express his exact query and different queries must be formulated to obtain his goals.

This work proposes a Pictorial Geographical Query Language, GeoPQL, that allows the user to represent only the desired relationships and avoid undesired relationships in the query’s visual representation.

The language is based on twelve operators. The set of operators includes all the main topological operators, distance and two operators devoted to solving ambiguities in visual query representation. The paper then discusses syntactic and semantic correctness of spatial configurations and related operators in the context of the declarative geographic pictorial query language. Some possible ambiguities and their solutions are presented in order to show the language’s characteristics.

GeoPQL has been implemented as a stand alone tool which interfaces with ESRI’s ArcView®, and the main results obtained are: high expressive power, solution of the ambiguities inherent to the spatial representation of a query and exact matching between the query and the obtained results.

Despite recent progress in the development of temporal Geographical Information Systems (GIS) there is still a lack of methodological integration with geophysical models oriented to the study of Earth changes. This paper introduces a temporal GIS modelling approach which complements a process-based geomorphological experimental apparatus that simulates erosion-sedimentation phenomena over a geological period of time. We combine a field-based with a discrete-based observation of forms and changes at different levels of abstraction. A fuzzy-based model of evolution is introduced and allows for an approximation of changes and processes. State transitions are fuzzy-valued and complemented by a quantitative analysis of change patterns.

Environmental planning requires constant tracing and revision of activities. Planners must be provided with appropriate documentation tools to aid communication among them and support plan enactment, revision and evolution. Moreover, planners often work in distinct institutions, thus these supporting tools must interoperate in distributed environments and in a semantically coherent fashion. Since semantics are strongly related to use, documentation also enhances the ways in which users can cooperate. The emergence of the Semantic Web created the need for documenting Web data and processes, using specific standards. This paper addresses this problem, for two issues: (1) ways of documenting planning processes, in three different aspects: was done, it was done and it was done that way; and (2) a framework that supports the management of those documents using Semantic Web standards.

A crucial issue during semantic integration of different geographic metadata sources is category comparison and reconciliation. We focus on the development of a framework for identification and resolution of semantic heterogeneity between geographic categories. The framework is divided in three processes: extraction, comparison and reconciliation. The first process performs semantic information extraction and formalization based on definitions of geographic category terms. Definitions constitute important sources of semantic information for geographic categories. Based on specific rules, definitions are analyzed in a set of semantic elements (properties and values). This information is further used in the second process to identify similarities and heterogeneities between geographic categories. Heterogeneity reconciliation is implemented by semantic factoring, a conceptual analysis process which results in a set of non-redundant, non-overlapping categories.

The interoperability problem arises in heterogeneous systems where different data sources coexist and there is a need for meaningful information sharing. One of the most representive realms of diversity of data representation is the spatio-temporal domain. Spatio-temporal data are most often described according to multiple and greatly diverse perceptions or viewpoints, using different terms and with heterogeneous levels of detail. Reconciling this heterogeneity to build a fully integrated database is known to be a complex and currently unresolved problem, and few formal approaches exist for the integration of spatio-temporal databases. The paper discusses the interoperation issue in the context of conceptual schema integration. Our proposal relies on two well-known formalisms: conceptual models and description logics. The MADS conceptual model with its multiple representation capabilities allows to fully describe semantics of the initial and integrated spatio-temporal schemas. Description logics are used to express the set of inter-schema mappings. Inference mechanisms of description logics allow us to check the compatibility of the semantic mappings and to propose different structural solutions for the integrated schema.

As location-based applications become part of our everyday life, ranging from traffic prediction systems to services over mobile phones providing us with information about our surroundings, the call for more semantics and accurate services is emerging. In this work, we analyze and register the data semantics of Location-based Services (LBS). Initially, we categorize LBS data according to the related concepts and use. We distinguish the (a) Domain Data, including spatial and temporal concepts, namely, position, location, movement and time, (b) Content Data, describing the LBS specific content, and (c) Application Data, consisting of the user profile and the services provided by LBS. Next, we model these three data categories in a way that captures their peculiarities and allows their sharing and exchange among different LBS, when desired. For this, we use semantically rich and expressive models, like UML, as well as the long-praised method of ontologies, realized in the open source, ontology and knowledge-based editor Protégé. To argue about the design choices and show their applicability, we present examples from two characteristic real-world applications, both in the Athens Metropolitan Area: an LBS for tourists carrying mobile devices, and a traffic LBS informing drivers about troublesome situations.

In 2001 David Mark and Barry Smith published a study that aimed to establish how non-expert subjects conceptualise geospatial phenomena in the United States of America. This paper contributes to comparing the results from the study performed by David Mark and Barry Smith with a similar study applied to Portuguese non-expert subjects.

In response to a series of questions, differently phrased, 160 non-expert subjects (university students from several parts of Portugal and various academic areas) were asked to give examples of geographical categories such as “Natural earth formation” or “Something that can be portrayed on a map”. The answers were used to statistically count the most mentioned terms. The Portuguese results were compared with those of the United States.

Although Portuguese results differ slightly in the number of items presented, the most mentioned terms match with the United States results.

We are therefore able to derive that the conceptualisation of geographical entities of Portuguese subjects is similar to that of United States subjects, e.g. Portuguese subjects also refer to mostly physical characteristics such as trees and mountains.

At a time when interoperability and ontological studies gain importance in the Geographical Information Systems/Science, this ongoing initiative points out the need of integrating trans-border geographical conceptualisations.