Re: Detecting cycles in OWL

Not so.

For example, how can you write SPARQL CONSTRUCT queries that can handle OWL
classification entailments such as

x:c owl:disjointUnionOf ( x:c1 [owl:complementOf x:c1] ) .

owl-entails

owl:Thing rdfs:subClassOf x:c .


Yes, yes, for any particular OWL derivation you could write a SPARQL CONSTRUCT
query that mimics that proof, but there are an infinite number of these
derivations.



peter



On 5/4/20 8:11 PM, Cox, Simon (L&W, Clayton) wrote:
>> IMO, SPARQL is far more general.
> Yeah. And pretty much any OWL entailment can be implemented as a SPARQL CONSTRUCT. 
>
> -----Original Message-----
> From: David Booth <david@dbooth.org> 
> Sent: Tuesday, 5 May, 2020 02:19
> To: semantic-web@w3.org
> Subject: Re: Detecting cycles in OWL
>
> On 5/4/20 11:16 AM, Mikael Pesonen wrote:
>> could be useful to have a
>> library of common reasoning cases with descriptions how to solve them 
>> and with what technology. For example this detecting of cycles as one 
>> case.
> That's a very reasonable suggestion, but IMO the real answer is that OWL is the wrong tool for that job.  OWL just isn't well suited to common graph analysis problems like shortest path or detecting cycles.
>
> I think RDF newcomers often misperceive OWL as a being a general tool that you should use, if you are using RDF.  But in my experience, OWL is much more of a niche technology than a general purpose tool.  It is good for a few use cases, but not most.  IMO, SPARQL is far more general. 
> But of course, this opinion is entirely a matter of use cases and personal preference, so your mileage may vary.
>
> David Booth
>
>
>
>

Received on Tuesday, 5 May 2020 00:54:41 UTC