Search Mailing List Archives


Limit search to: Subject & Body Subject Author
Sort by: Reverse Sort
Limit to: All This Week Last Week This Month Last Month
Select Date Range     through    

[protege-owl] Problems with Manchester OWL Editor

Timothy Redmond tredmond at stanford.edu
Sat Nov 20 15:54:43 PST 2010


> Correct me if I am wrong, but I didn't see this problem with Protege 4.1
> You are wrong.

You are right - I was wrong.  I was very sloppy in checking things out.

I think that what is needed here is better validation tools.  This 
particular item would be very easy to discover programatically.   I 
suspect the OWL api OWL 2 validator would find it.  Integrating this 
type of validation with an ontology editor is a bit less trivial but it 
is definitely in the plans.   There is a need for better validation 
because we get a fair bit of e-mail on the list with people having 
trouble with the global property hierarchy constraints.  Also we are 
having periodic problems with ontologies that do not parse under 1.0 or 
2.0 because they use 1.1 rdf idioms.

But I think it is good that the tools try to load and work with 
problematic ontologies.  We are seeing such ontologies and people want 
to work with them.  In some cases the ontology developers are explicitly 
resistant to making these ontologies OWL 2 compliant.  Logging a warning 
is nice but is often not particularly helpful.  You could also use 
layered error levels and handlers (ala Jena and Protege 3 OWL) but this 
is a bit ugly.

I am curious how rdfs:label became an object property.  This is not 
clear yet.

-Timothy




On 11/20/2010 11:54 AM, Alan Ruttenberg wrote:
> On Sat, Nov 20, 2010 at 11:40 AM, Timothy Redmond<tredmond at stanford.edu>  wrote:
>> All right. I see what you see (you need to update to protege 4.1, btw
>> - 4.02 is obsolete)
>>
>> This is right.
>>
>> I'm ccing this to Matthew Horridge, who writes the OWLAPI, and to the
>> reasoner developers as well. Hopefully between them and Tim they will
>> figure out what's broken where and repair this.
>>
>> Correct me if I am wrong, but I didn't see this problem with Protege 4.1
> You are wrong. As usual, I test with the latest version before I report a bug.
> The file attached to my previous message, which is invalid OWL-DL
> loads and reasons without note.
>
> You can see the issue by looking for the name property, which you will
> see in both the object property(subproperty of label) and data
> property (top level) tabs. It has the same URI in both cases -
> http://xmlns.com/foaf/0.1/name
>
> In the individuals tab you can see an individual named "twonames"
> which uses the property as both an object and data property.
>
> For reference, the relevant section of the spec is
> http://www.w3.org/TR/2009/REC-owl2-syntax-20091027/#Typing_Constraints_of_OWL_2_DL
> "No IRI I is declared in Ax as being of more than one type of
> property; that is, no I is declared in Ax to be both object and data,
> object and annotation, or data and annotation property."
>
> -Alan
>
>> (and thus with the 3.1 version of the OWL api).  It doesn't make sense at
>> this point to fix OWL API 3.0 "bugs".  This api targeted a specification of
>> the OWL language (1.1) that never became a standard.  In fact, I think the
>> real story is even more muddy than that.
>>
>> So in order to have an issue that needs fixing, it needs to be demonstrated
>> to be an OWL api 3.1 issue.
>>
>> -Timothy
>>
>>
>>
>> On 11/20/2010 06:12 AM, Alan Ruttenberg wrote:
>>
>> All right. I see what you see (you need to update to protege 4.1, btw
>> - 4.02 is obsolete)
>> This issue exposes several bugs.
>>
>> Here is the offending clause:
>>
>>    <rdf:Property rdf:about="http://xmlns.com/foaf/0.1/name"
>> vs:term_status="testing" rdfs:label="name" rdfs:comment="A name for
>> some thing.">
>>      <rdf:type
>> rdf:resource="http://www.w3.org/2002/07/owl#DatatypeProperty"/>
>>      <rdfs:domain rdf:resource="http://www.w3.org/2002/07/owl#Thing"/>
>>      <rdfs:range
>> rdf:resource="http://www.w3.org/2000/01/rdf-schema#Literal"/>
>>      <rdfs:isDefinedBy rdf:resource="http://xmlns.com/foaf/0.1/"/>
>>      <rdfs:subPropertyOf
>> rdf:resource="http://www.w3.org/2000/01/rdf-schema#label"/>
>>    </rdf:Property>
>>
>> 1) The foaf file is not valid OWL-DL (nor does it claim to be). In
>> OWL, rdfs:label is an annotation property. As name is a datatype
>> property. Datatype properties can't be subproperty of annotation
>> property. The tool is not telling you that.
>>
>> 2) Even though rdfs:label is nowhere declared to be an object
>> property, protege/owlapi is deciding it is one.
>>
>> 3) Having decided it is one, since name is a subproperty of label, it
>> is deciding it too is an object property.
>>
>> 4) Properties can't be both object and data properties in OWL DL. The
>> tool is not telling you that.
>>
>> 5) Worse, you can actually use the property as both a data property
>> and an object property and none of the reasoners will complain.
>> (attached file).
>>
>> I'm ccing this to Matthew Horridge, who writes the OWLAPI, and to the
>> reasoner developers as well. Hopefully between them and Tim they will
>> figure out what's broken where and repair this.
>>
>> Unfortunately, if you want to use foaf in OWL DL, then you must either
>> petition the developers to make one that is OWL DL compatible (not
>> hard - I could help) or simply selected terms you want to use within
>> protege. You can set the URI of a term using the Refactor>Change
>> entity URI after you define a property to set its URI to the foaf URI.
>>
>> Tim, the Gforge item isn't relevant here, as this isn't a valid case of
>> punning.
>>
>> -Alan
>>
>>
>> On Sat, Nov 20, 2010 at 3:18 AM, Guillermo Garcia
>> <guillermo at ggarciao.com>  wrote:
>>
>> Sure,
>> * Im using protege 4.0.2 - binary version with installer for windows.
>> * The FOAF ontology uri is: http://xmlns.com/foaf/0.1/
>>
>> On Fri, Nov 19, 2010 at 4:15 PM, Alan Ruttenberg<alanruttenberg at gmail.com>
>> wrote:
>>
>> On Fri, Nov 19, 2010 at 2:17 PM, Guillermo Garcia
>> <guillermo at ggarciao.com>  wrote:
>>
>> I saw it in the protege GUI:
>> * name is a Data Property with range Literal
>> * name is also a Object Property, decendent of label Object Property
>>
>> Can you send me a link to the version you imported? Or the ontology as a
>> whole?
>> Are you using Protege 3 or Protege 4?
>> -Alan
>>
>> On Fri, Nov 19, 2010 at 1:32 PM, Alan Ruttenberg
>> <alanruttenberg at gmail.com>
>> wrote:
>>
>> On Fri, Nov 19, 2010 at 6:21 AM, Guillermo Garcia
>> <guillermo at ggarciao.com>  wrote:
>>
>> Person and name value "Peter Parker"
>> where
>>     Person ->  FOAF ontology Class
>>     name ->  FOAF ontology Data Property
>>     "Peter Parker" ->  Literal
>> But the query editor doesnt let me execute the query because he cant
>> resolve
>> the conflict between the "name" Data property and the "name" Object
>> Property
>> (both define en FOAF ontology).
>>
>> Where are these both defined in foaf? The documentation says the range
>> is a literal. In the RDF I see:
>>
>>   <rdf:Property rdf:about="http://xmlns.com/foaf/0.1/name"
>> vs:term_status="testing" rdfs:label="name" rdfs:comment="A name for
>> some thing.">
>>     <rdf:type
>> rdf:resource="http://www.w3.org/2002/07/owl#DatatypeProperty"/>
>>     <rdfs:domain rdf:resource="http://www.w3.org/2002/07/owl#Thing"/>
>>     <rdfs:range
>> rdf:resource="http://www.w3.org/2000/01/rdf-schema#Literal"/>
>>     <rdfs:isDefinedBy rdf:resource="http://xmlns.com/foaf/0.1/"/>
>>     <rdfs:subPropertyOf
>> rdf:resource="http://www.w3.org/2000/01/rdf-schema#label"/>
>>   </rdf:Property>
>>
>> -Alan
>>
>> The Editor always think that name is the
>>
>> Object Property and due to that, I cant put a Literal as part of the
>> query.
>> This is a name scope bug? or there is a way to solve this using the
>> syntax
>> of the query language (like prefixs, for exemple)?
>>
>> --
>> Guillermo Tomás García Ochoa
>>          www.ggarciao.com
>>
>>
>> _______________________________________________
>> protege-owl mailing list
>> protege-owl at lists.stanford.edu
>> https://mailman.stanford.edu/mailman/listinfo/protege-owl
>>
>> Instructions for unsubscribing:
>> http://protege.stanford.edu/doc/faq.html#01a.03
>>
>>
>> _______________________________________________
>> protege-owl mailing list
>> protege-owl at lists.stanford.edu
>> https://mailman.stanford.edu/mailman/listinfo/protege-owl
>>
>> Instructions for unsubscribing:
>> http://protege.stanford.edu/doc/faq.html#01a.03
>>
>>
>> --
>> Guillermo Tomás García Ochoa
>>     Portable: (+33) 6.59.80.16.78
>>          www.ggarciao.com
>>
>>
>> _______________________________________________
>> protege-owl mailing list
>> protege-owl at lists.stanford.edu
>> https://mailman.stanford.edu/mailman/listinfo/protege-owl
>>
>> Instructions for unsubscribing:
>> http://protege.stanford.edu/doc/faq.html#01a.03
>>
>>
>> _______________________________________________
>> protege-owl mailing list
>> protege-owl at lists.stanford.edu
>> https://mailman.stanford.edu/mailman/listinfo/protege-owl
>>
>> Instructions for unsubscribing:
>> http://protege.stanford.edu/doc/faq.html#01a.03
>>
>>
>> --
>> Guillermo Tomás García Ochoa
>>     Portable: (+33) 6.59.80.16.78
>>          www.ggarciao.com
>>
>>
>> _______________________________________________
>> protege-owl mailing list
>> protege-owl at lists.stanford.edu
>> https://mailman.stanford.edu/mailman/listinfo/protege-owl
>>
>> Instructions for unsubscribing:
>> http://protege.stanford.edu/doc/faq.html#01a.03
>>
>>
>>
>> _______________________________________________
>> protege-owl mailing list
>> protege-owl at lists.stanford.edu
>> https://mailman.stanford.edu/mailman/listinfo/protege-owl
>>
>> Instructions for unsubscribing:
>> http://protege.stanford.edu/doc/faq.html#01a.03
>>
>>
>> _______________________________________________
>> protege-owl mailing list
>> protege-owl at lists.stanford.edu
>> https://mailman.stanford.edu/mailman/listinfo/protege-owl
>>
>> Instructions for unsubscribing:
>> http://protege.stanford.edu/doc/faq.html#01a.03
>>
>>
> _______________________________________________
> protege-owl mailing list
> protege-owl at lists.stanford.edu
> https://mailman.stanford.edu/mailman/listinfo/protege-owl
>
> Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03




More information about the protege-owl mailing list