Nada

^ Up to Research, Theory group at Nada, KTH.

Object-Oriented Spatial Analysis DBMS

Researcher

Ph.D. student

Short description

This project researches and evaluates state-of-the-art spatial database technology with a view towards improving the interface between users and Geographical Information Systems (GIS), as well as between GIS and databases.

Current GIS's lack efficient facilities for integrating the base system and a domain model for spatial analysis. Such facilities may be provided by an object-oriented database language, whose primitives represent the fundamental operations of spatial analysis. The use of declarative object-oriented queries for domain modelling offers advantages over conventional imperative programming techniques:

  • object-oriented query languages permit advanced ad-hoc queries;
  • declarative models are easier to define, inspect and understand than imperative programs and lead to compact, easily reusable, and thus powerful domain models;
  • object-oriented query languages provide object views which are invoked independently of whether they represent stored or derived data. This feature supports data independence and schema evolution;
  • the extensibility feature of certain "next-generation" ODBMS's permit optimization of expressions involving user defined operations.

    The project has the following objectives:

  • Develop a prototype spatial analysis and database management system based on the AMOS main-memory object database handler (Risch & al, Linkoeping University) with spatial extensions, functionally similar to the GeoSAL proposal (Svensson & Huang, KTH 1990-1993);
  • For the spatial extensions, spatial operator implementations from the ROSE library (Gueting & al, Fernuniversitaet Hagen 1993-) will be used;
  • Evaluate the prototype over a representative set of spatial data analysis tasks.

    The ultimate goal of the project is to show that proper application of next-generation ODBMS extensibility concepts to GIS can provide a generic environment in which a large class of realistic and complex spatial data analysis problems can be defined and solved with previously unattainable clarity and ease.

    Funding

    The project is funded 1996-1998 by Centre for Geoinformatics.

    Recent publications

    AMORose, a Realm-based Spatial Database System
    K. Oukbir
    Scandinavian Conference on GIS 1997 (ScanGIS 97)

    ^ Up to Research, Theory group at Nada, KTH.

    Responsible for this page: Viggo Kann <viggo@nada.kth.se>
    Latest change October 8, 1997
    Technical support: <webmaster@nada.kth.se>