CEDEX

CEDEX an extensible ontologie for the transfer of ecological data

Schentz Herbert , Mirtl Michael , Umweltbundesamt GmbH , Austria

Challenge

The Austrian environment agency has created the object-relational Informationsystem MORIS to manage all the data generated by the Austrian part of the UN ECE program “Integrated Monitoring”. The concept of this information system with its core and rules for extension proofed to fit for (nearly) any ecological topic.

Now as the agency would strongly need an universal interface for the exchange of ecological data, we supposed the concept would also be appropriate for that task. We choose an object orientated formal XML based language, OWL, on which we mapped the basic structure and the examples for extensions. We called the result CEDEX (Classes for Environmental Data Exchange)

Concept of CEDEX

We fixed right from the beginning:

  • The definition has to be object orientated which could easily be done, since MORIS already is objectorientated.
  • The base ontology shall be easily extended, where the extensions must be as independent from each other as possible.
  • The structure must take into account, that Paramters, Methodes and so on may be related rather in the same way as taxons are.

Realisation

(1)There is a very simple base ontology.

(2)All extension classes are derived from those base classes. No independent new classes may be built.

(3)All existing relations are defined in the base classes.The relations of the extension classes may just restrict the allowed classes.

The core of CEDEX-BASE is a rearrangement of the well known "geographical coverage, temporal coverage, taxonomic coverage" (© EML et al.):

  • OBJEKT and derived classes register all the things, on which observations are done. (And also all auxiliary structures)
  • PARAMETER and derived classes contain all the things that are observed.
  • METHODE and derived classes are the methods of observation.
  • The association class DATENPUNKT notifies where which parameters are observed.
  • PROZESS and derived classes are for natural processes (e.g. meteorological and geomorphological events) and human made processes (testings, ….)

fig 1 shows the basic classes (white) and the first generation of derived classes without attributes. It goes without saying that a tree of inherited classes is allowed.

Some auxiliary classes (like scaling, dimension, …..) and some attachment classes are defined. The attachment classes may be extended in the same way as the basic classes.

We added some examples for extensions.

Fig. 2 shows CEDEX BASE and some extensions.

As one can already see in fig. 1 a relation between the objects derived from one basic object is defined. Taxonomists may well know this relationship, which allows the mapping of a polyhierarchy. We know that taxonomists are very astonished when they are told that not only species but also parameters like “heavy metal” are not always the same, depending on methods of gathering, preparing the specimen, accurancy and so on of measurement.

CEDEX allows the definition of very simple functions that can be applied according to the relation “polyhierarchy”. This is shown in the UML schema.

Downloads

  • UML definition of CEDEX
  • Protégé 2.0 definition of CEDEX
  • OWL definition of CEDEX

links

OWL Web Ontology Language:

Science Environment for Ecological Knowledge (SEEK):

Protégé:

Ecological Metadata Language (EML):

Schentz12004-03-09