Re: Repository creation problem

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

Re: Repository creation problem

Matthew Horridge
Hi Daniel,

>> The problems that you are experiencing are due to the way that the  
>> repository mechanism works.  The repository manager uses the  
>> heuristic that the URI of an ontology is the xml base of the  
>> document that the ontology is defined in - if there is no xml base  
>> then the URI is taken as the URL of the document containing the  
>> ontology.  Unfortunately, the dc terms ontology does not have an  
>> xml base so this mechanism fails for local imports.
>>
> Why don't you use the heuristic:
> 1) If there is an Ontology element, use its URI,

Well there could be several ontology elements in an RDF/XML document  
- it's difficult to know which is the "right" one.

> 2) Otherwise, if there is an xml:base, use it as the ontology URI,

This is what I currently do - we assume the ontology will be  
published at a location that corresponds to the xml:base

> 3) Otherwise, if there is a default namespace, use it as the  
> ontology URI

This is another possibility

Thanks for the suggestions - I will consider them.

Cheers,

Matthew
-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html

Reply | Threaded
Open this post in threaded view
|

Re: Repository creation problem

Daniel Elenius
Matthew Horridge wrote:

> Hi Daniel,
>
>>> The problems that you are experiencing are due to the way that the
>>> repository mechanism works.  The repository manager uses the
>>> heuristic that the URI of an ontology is the xml base of the
>>> document that the ontology is defined in - if there is no xml base
>>> then the URI is taken as the URL of the document containing the
>>> ontology.  Unfortunately, the dc terms ontology does not have an xml
>>> base so this mechanism fails for local imports.
>>>
>> Why don't you use the heuristic:
>> 1) If there is an Ontology element, use its URI,
>
> Well there could be several ontology elements in an RDF/XML document -
> it's difficult to know which is the "right" one.
>
Oh! I never considered this possibility. Does the OWL spec even allow
for this? It doesn't make sense to me. How do you know which ontology
the rest of the stuff in the file belongs to? Does Protege support this?

>> 2) Otherwise, if there is an xml:base, use it as the ontology URI,
>
> This is what I currently do - we assume the ontology will be published
> at a location that corresponds to the xml:base
>
>> 3) Otherwise, if there is a default namespace, use it as the ontology
>> URI
>
> This is another possibility
>
> Thanks for the suggestions - I will consider them.
>
> Cheers,
>
> Matthew-------------------------------------------------------------------------
>
> To unsubscribe go to http://protege.stanford.edu/community/subscribe.html

-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html

Reply | Threaded
Open this post in threaded view
|

Re: Repository creation problem

Matthew Horridge

On 20 Jan 2006, at 18:00, Daniel Elenius wrote:

> Matthew Horridge wrote:
>> Hi Daniel,
>>
>>>> The problems that you are experiencing are due to the way that  
>>>> the repository mechanism works.  The repository manager uses the  
>>>> heuristic that the URI of an ontology is the xml base of the  
>>>> document that the ontology is defined in - if there is no xml  
>>>> base then the URI is taken as the URL of the document containing  
>>>> the ontology.  Unfortunately, the dc terms ontology does not  
>>>> have an xml base so this mechanism fails for local imports.
>>>>
>>> Why don't you use the heuristic:
>>> 1) If there is an Ontology element, use its URI,
>>
>> Well there could be several ontology elements in an RDF/XML  
>> document - it's difficult to know which is the "right" one.
>>
> Oh! I never considered this possibility. Does the OWL spec even  
> allow for this?

Yes - it does allow this

>  It doesn't make sense to me.

I suppose you can say anything about any resource, including  
resources which are typed as ontologies.

> How do you know which ontology the rest of the stuff in the file  
> belongs to?

It doesn't belong to any ontology - I have reliably been informed  
that there is no link between between the statements an RDF/XML  
document and the resource which is typed as an ontology.

> Does Protege support this?

You can do this with the API. It's not obvious how to do it through  
the UI - although you can :)

Cheers,

Matthew


>>> 2) Otherwise, if there is an xml:base, use it as the ontology URI,
>>
>> This is what I currently do - we assume the ontology will be  
>> published at a location that corresponds to the xml:base
>>
>>> 3) Otherwise, if there is a default namespace, use it as the  
>>> ontology URI
>>
>> This is another possibility
>>
>> Thanks for the suggestions - I will consider them.
>>
>> Cheers,
>>
>> Matthew--------------------------------------------------------------
>> -----------
>> To unsubscribe go to http://protege.stanford.edu/community/ 
>> subscribe.html
>
> ----------------------------------------------------------------------
> ---
> To unsubscribe go to http://protege.stanford.edu/community/ 
> subscribe.html
>

-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html