Détail de l'auteur
Auteur J. Althoff |
Documents disponibles écrits par cet auteur (1)
![](./images/expand_all.gif)
![](./images/collapse_all.gif)
![Tris disponibles](./images/orderby_az.gif)
Titre : Model-driven tools to support conceptual geospatial modelling Type de document : Monographie Auteurs : J. Althoff, Auteur Editeur : Zurich : Eidgenossische Technische Hochschule ETH - Ecole Polytechnique Fédérale de Zurich EPFZ Année de publication : 2011 Collection : Dissertationen ETH num. 19918 Importance : 193 p. Format : 21 x 30 cm Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Géomatique
[Termes IGN] architecture orientée modèle
[Termes IGN] GML
[Termes IGN] INTERLIS
[Termes IGN] langage de modélisation
[Termes IGN] MADS
[Termes IGN] métamodèle
[Termes IGN] modèle conceptuel de données localisées
[Termes IGN] UMLRésumé : (Auteur) The raising importance and fast development of national and international Spatial Data Infrastructures originate a need for harmonised geo data of different sources. Spatial datasets should not only be found, presented and processed via geo portals, but different information should also be capable of being integrated. Beyond pure technical and geometric issues like file formats, spatial resolutions or reference systems, this also includes informational aspects as data structures and semantic aspects. A precondition for a tight integration of spatial data from different sources is precise definition of the used or targeted datasets. Such definition processes currently take place in the 'INSPIRE data specifiations' or the Swiss efforts to create 'minimum data models' according to the Swiss federal act on geoinformation. In those initiatives groups of experts discuss about common information elements, how those elements are characterised in detail and relate to each other. The resulting data description is formalised in so called 'conceptual data models', which should be as independent of data formats, database systems or general computational aspects as possible. Data models in the field of spatial information are characterised by specific aspects that are not common in other domains. Those are especially location with reference systems, different kinds of 2d or 3d geometries or measurements with units and value ranges. With common 'general purpose' modelling approaches like UML it is difficult to represent those aspects in a sufficient and user friendly way. Consequently, in recent years several specialised approaches were developed to support the creation of spatial models. There are extensions to UML, like GeoUML or the ISO/TC-211 standards, as well as independent languages like INTERLIS or MADS. Drawbacks of all specialisations are mainly a lack of suitable tools for modelling, limitations in scope or adaptability and poor interoperability between different approaches. Such issues are addressed by new trends in the field of model driven software developments. After monolithic and inflexible approaches, new 'language-centric' methods evolve to support the adaptability of modelling languages, the creation of suitable tools and transformations between different types of models. A core technology for this is 'meta modelling' and describes the formal definition of modelling languages with standardised elements. Meta models of modelling languages, which is used to efficiently generate software tools like graphical or textual editors, model validators or model transformations. This approach is highly efficient and not only used for existing languages, but also induced the creation of highly specialised 'small' (or 'domain-specific') languages, which are tightly adjusted to a field of work, a specific task or a user group. This thesis explores to what extend those ideas of model driven software development can be applied to conceptual spatial data modelling. This could help to simplify modelling for domains expert by providing them suitable languages and graphical editors, enhance model quality with model validators and offer new ways of interoperability between different modelling languages. Firstly it will be tested, if a common meta modelling approaches of software development can be used to sufficiently describe languages for conceptual spatial modelling. This is carried out for existing conceptual geospatial languages with Ecore, the most commonly used meta modelling approach. As result, the different languages can be represented on a common definition base. Additionally to already existing modelling languages, the idea of 'domain specific languages' is taken up to develop an exemplary modelling language 'HML' for the HUMBOLDT project1. According tools like a graphical model editor and a validating parser are created in a highly automated way for this language. Based on the developed meta models, transformations between different conceptual spatial modelling approaches are investigated in a second step. A language-to-language mapping between the domain specific approach of HML and ISO/TC-211 compliant UML is declared and executed. Such transformations between specialised languages and general purpose languages are of particular interest, because they offer ways of optimising modelling processes while staying interoperable to existing models and tools. Finally vertical transformations, the possibility to create logical or physical models from the conceptual models, are addressed. Similar to the horizontal language-to-language transformations, standard tools from software development are used to derive GML Schema and textual model documentation. All in all, this work proves along practical implementations how language based technologies of modern model driven software development can support conceptual spatial modelling. Especially the combination of 'domain-specific' modelling languages with according software tools and horizontal language-to-language tools offers possibilities to simplify modelling processes, increases model quality and opens a new field of interoperability between different conceptual modelling languages. Note de contenu : 1 Introduction
1.1 Motivation and Current Situation
1.2 Hypothesis
1.3 Structure of the Thesis
1.4 Summary
2 State of the Art
2.1 Modelling and Models
2.2 Conceptual Modelling Languages
2.3 Model Transformation
2.4 Model-Driven Software Development
2.5 Summary
3 Concepts
3.1 Problem Field
3.2 Working Thesis
3.3 Meta Models
3.4 Horizontal Transformations
3.5 Vertical Transformations
3.6 Summary
4 Implementation
4.1 Used Technology
4.2 Domain Specific Language & Graphical Editor
4.3 Language-to-Language Transformations
4.4 Realisation of Vertical Transformations
4.5 Summary
5 Summary of the Results
6 Discussion
6.1 Definition of Domain Specific Languages
6.2 Tools
6.3 Horizontal Language-to-Language Transformations
6.4 OutlookNuméro de notice : 14643 Affiliation des auteurs : non IGN Thématique : GEOMATIQUE Nature : Monographie Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=62676 Documents numériques
en open access
14643_dissertationeth-4746-02_althoff.pdfAdobe Acrobat PDF