Need to be able to add imports to ontologies

[GoogleCodeExporter]

You need to be able to specify imports on an ontology, and for the "superclass" 
window to display the classes in those imports as candidate superclasses when 
you create a new class.

Original issue reported on code.google.com by pmurray....@gmail.com on 13 May 2011 at 9:12

[GoogleCodeExporter]

owl:imports is rarely used in web vocabularies, therefore it's not supported in 
Neologism.

You can add an “External Vocabulary” to make classes from other 
vocabularies/ontologies available as candidate superclasses.

Original comment by richard....@gmail.com on 13 May 2011 at 1:00

[GoogleCodeExporter]

This also is a surprising response.

As I mentioned against my other issue, I am working with the bioinformatics 
community. We need a tool to permit subject-matter experts (scientists) to 
build vocabularies. I think that we have three basic use-cases for ontologies.

1) the creation of very general ontologies over the field, describing concepts 
common to all of biological taxonomy (taxon, organism, scientific name, etc)

2) creation of vocabularies for specific subfields.

3) creation of vocabularies whose purpose it is to describe data made available 
from some database - the classes an properties more-or-less corrwsponding to 
database tables and columns.

Generally speaking, to be of any use at all the more specific ontologies need 
to declare equivalencies with the more general ones. In case 2) above, the 
herpetologists might want to greate an ontology for types of snake scale. 
That's all fine and dandy, but the real power of the semantic web is that they 
can also say "snake-scale-type is something that it's meaningful to say about a 
specimen. It's a particular kind of *description*", where description is a term 
in the more general ontology.

Likewise in case 3, the real power of the semantic web is being able to say 
"column SCALE_TYP in table DESCR" (ie: predivate SCALE_TYP with a domain of 
DESCR) "is a subproperty of Prof. Foo's herpetology ontology".

Once you do that, then you can ask sparql question like "what datasets have any 
data relating to scales? Or teeth? Or location?"

Making a linked world of vocabularies each defined by their various experts is 
the whole point of the exercise. and it involves imports - being able to say to 
sparql (or any other reasoner) "you cannot make sense of my terms without also 
pulling down this other vocabulary".

Original comment by pmurray....@gmail.com on 14 May 2011 at 12:07

[GoogleCodeExporter]

You appear to be looking for a full-featured OWL ontology editor. There are 
plenty of those around. Neologism might well be the wrong product for you. It 
doesn't aim at OWL but at RDFS. It borrows some constructs from OWL, but makes 
no claim whatsoever at being OWL-compatible in any way.

It does not use owl:imports because that construct doesn't make any sense if 
one takes the Web seriously. We name concepts with URIs. Resolvable URIs. If 
one needs a definition of the concept, one resolves the URI and gets an 
authoritative definition from the URI owner. There's no need for owl:imports. 
On the Web, owl:imports is as useful as an index in a dictionary.

Original comment by richard....@gmail.com on 15 May 2011 at 3:10

[GoogleCodeExporter]

I also should mention the “Additional custom RDF” feature, where you can 
add arbitrary additional RDF statements to a vocabulary, using Turtle syntax. 
Plopping some owl:imports statements in there is easy. They will not show up in 
the HTML though.

Original comment by richard....@gmail.com on 15 May 2011 at 3:27