I like Class and Property too. Some problems:
1. We have to define these things! To this end I have started a wiki page (http://wiki.tdwg.org/twiki/bin/view/TAG/TDWGOntologyTerminology). With the start of some tentative definitions on. It would be good to have in brackets what the term might equate to in different technologies (UML, RDFS, OWL, XSD, SQL etc) or if it doesn't match anything exactly. Remember we are defining data models not software artifacts so we can't just pinch the Java definition of a class for example. 2. We still have the problem of the use of 'Concept' in the mapping of BioCASE and TAPIR protocols. I don't see that this term maps directly to either Class or Property. It is equates to an XML Schema element or attribute in current usage. Do we need a new name for it or a way of qualifying it's use, 'mapping-concept' perhaps? 3. What about relationships? I see these as just properties of a class with the range restricted to objects of a class. This keeps it really simple and maps well to SQL (foreign keys) and RDFS (property with domain of one class and range of another). I'd rather not get into bidirectional relationships (would need OWL express this) at the beginning though we may want to add them later.
Please lets keep the discussion on the mailing list and the results of discussion on the wiki where possible. Not everyone follows the wiki and the wiki becomes unreadable a documentation if it contains long of dialogues. It is always possible to link to the thread in the email discussion archives (http://lists.tdwg.org/pipermail/tdwg-tag_lists.tdwg.org/) from the wiki and /vice versa/
All the best,
Roger
Gregor Hagedorn wrote:
Class and property is ok, although these are frequently overloaded as well, they are well established in software design.
See also second email.
Gregor
I second the choice of class and properties. With this terminology Darwin Core is (currently) a class of a specimen record and each of what we once called concepts therein is a property. Of course, we will likely have to use alternate terminology to talk to most of the rest of the world (provider representatives), who will not understand what we mean by "class" and "property", but would more likely understand "table" and "field". Still, it's better for us to be internally consistent.
On 5/2/06, Robert Gales rgales@ku.edu wrote:
I agree, concept is far too overloaded a term to use for general conversations regarding ontologies. Not only is the term overloaded across domains, even within a single domain, such as taxonomy, it is difficult at best to get consensus on the definition of concept.
I tend to prefer the terms class and properties. Whatever terms we decide on, we should probably have very brief definitions of them. That probably goes for any terminology we decide on within the group, not only so we understand one another, but outside readers understand any of the documentation that we produce.
- Rob
Roger Hyam wrote:
Hi All,
Gregor and I are having a discussion over what 'concept' means.
Gregor said: Roger, I suggest, you replace unqualified "concept" with something that qualifies what kind of concept you mean. If I read concept I think of a taxon concept or descriptive concept (flower, color, blue, frequently) but not of the kind of concepts you mean (I would say "data elements" for these, but I leave it to you. This was the source of the major misunderstanding about Management system and data. Note that both taxon and descriptive concepts have and need ontology information, even if not expressed in RDF - so just saying "ontology concepts" would not be very clear I believe.
I know TAPIR and BioCASE use concepts to mean element or attribute and believe DwC do as well. I thought I was on clear ground here in using it in the context of ontologies.
What word other than 'concept' can I use to describe a thing in an ontology i.e. a class, property, attribute whatever.
Any suggestions,
Roger
--
Roger Hyam Technical Architect Taxonomic Databases Working Group
http://www.tdwg.org roger@tdwg.org mailto:roger@tdwg.org
+44 1578 722782
Tdwg-tag mailing list Tdwg-tag@lists.tdwg.org http://lists.tdwg.org/mailman/listinfo/tdwg-tag_lists.tdwg.org
Tdwg-tag mailing list Tdwg-tag@lists.tdwg.org http://lists.tdwg.org/mailman/listinfo/tdwg-tag_lists.tdwg.org
Gregor Hagedorn (G.Hagedorn@bba.de) Institute for Plant Virology, Microbiology, and Biosafety Federal Research Center for Agriculture and Forestry (BBA) Königin-Luise-Str. 19 Tel: +49-30-8304-2220 14195 Berlin, Germany Fax: +49-30-8304-2203
Tdwg-tag mailing list Tdwg-tag@lists.tdwg.org http://lists.tdwg.org/mailman/listinfo/tdwg-tag_lists.tdwg.org