domains and ranges for object properties?
Egon Willighagen
I was looking at the object properties, and noted that many do not
have domains and ranges defined, such as 'has_attribute'... is that
deliberate? Are those added axiomatically? If so, how can I get the
reasoner to do that? I tried reasoning with HermiT, but that did not
add such information, so it does not seem to be inherited from
somewhere...
Egon
--
Dr E.L. Willighagen
Postdoctoral Research Associate
University of Cambridge
Homepage: http://egonw.github.com/
LinkedIn: http://se.linkedin.com/in/egonw
Blog: http://chem-bla-ics.blogspot.com/
PubList: http://www.citeulike.org/user/egonw/tag/papers
Janna Hastings
It's not deliberate, just an omission... You can add them as per the paper if you're busy there anyway? If not I will do it in a few days.
Cheers, Janna
Michel Dumontier
m.
--
Michel Dumontier
Associate Professor of Bioinformatics
Carleton University
http://dumontierlab.com
Egon Willighagen
<michel.d...@gmail.com> wrote:
> What do you propose?
Something like this:
CHEMINF_000200 (has attribute):
domain: CHEMINF_000000 (chemical entity)
range: IAO_000028 (data item)
CHEMINF_000143 (is descriptor of)
domain: CHEMINF_000123 (chemical descriptor)
range: specifically_dependent_continuant (this one is already axiomized)
Michel_Dumontier
'has attribute' is the domain of any entity (owl:Thing), but yes, 'data item' seems appropriate as a range.
'is descriptor of' would probably also have 'data item' as a domain, rather than the more restrictive 'chemical descriptor'
m.
Egon Willighagen
<Michel_D...@carleton.ca> wrote:
> Good. I counter propose:
Most welcome! Mine was just an initial guess, based on what the paper writes...
> 'has attribute' is the domain of any entity (owl:Thing), but yes, 'data item' seems appropriate as a range.
For example, referring to 'data item' is actually as listed in the draft...
> 'is descriptor of' would probably also have 'data item' as a domain, rather than the more restrictive 'chemical descriptor'
The manuscript should then be updated accordingly, and perhaps labels
too... I'll get back on this. I still need to look at other properties
too...
Michel Dumontier
<egon.wil...@gmail.com> wrote:
> On Thu, Nov 4, 2010 at 3:38 PM, Michel_Dumontier
> <Michel_D...@carleton.ca> wrote:
>> Good. I counter propose:
>
> Most welcome! Mine was just an initial guess, based on what the paper writes...
>
>> 'has attribute' is the domain of any entity (owl:Thing), but yes, 'data item' seems appropriate as a range.
>
> For example, referring to 'data item' is actually as listed in the draft...
>
>> 'is descriptor of' would probably also have 'data item' as a domain, rather than the more restrictive 'chemical descriptor'
>
> The manuscript should then be updated accordingly, and perhaps labels
> too... I'll get back on this. I still need to look at other properties
> too...
many thanks Egon!
m.
>
> Egon
>
> --
> Dr E.L. Willighagen
> Postdoctoral Research Associate
> University of Cambridge
> Homepage: http://egonw.github.com/
> LinkedIn: http://se.linkedin.com/in/egonw
> Blog: http://chem-bla-ics.blogspot.com/
> PubList: http://www.citeulike.org/user/egonw/tag/papers
>
--