Modelling the history of semi-structured geographical entities

Geographical entities are characterized by rather complex structures. They involve space and thematic information, which is subject to change in time, while history should be maintained. On the other hand, these structures may be irregular (i.e. they do not necessarily conform to a fixed schema), be...

Full description

Saved in:
Bibliographic Details
Published inInternational journal of geographical information science : IJGIS Vol. 17; no. 6; pp. 517 - 546
Main Author Stefanakis, Emmanuel
Format Journal Article
LanguageEnglish
Published Taylor & Francis Group 01.09.2003
Online AccessGet full text
ISSN1365-8816
1362-3087
DOI10.1080/1365881031000114080

Cover

More Information
Summary:Geographical entities are characterized by rather complex structures. They involve space and thematic information, which is subject to change in time, while history should be maintained. On the other hand, these structures may be irregular (i.e. they do not necessarily conform to a fixed schema), because associated data is usually collected based on different specifications and multiple resolutions. Hence, the representation of geographical entities in traditional data models, such as the relational or object-oriented, is not always feasible. In this respect, this paper investigates the use of semi-structured data (SSD) models-an innovative approach recently developed in Information Technology-for modelling dynamic geographical entities. A framework for the representation of geographic entities in Object Exchange Model (OEM), a popular model for semi-structured data, is introduced. Additionally, it is shown how useful information can be extracted from such a representation using the LOREL query language for SSD. A simplified case study in the application domain of cadastre involving SSD is examined closely.
ISSN:1365-8816
1362-3087
DOI:10.1080/1365881031000114080