Re: sameAs.org milestone

Hi Anna,
Thanks for your interest.
On 9 Feb 2012, at 10:59, Jordanous, Anna wrote:

> A very useful resource, which touches on a current issue for me - I may be reviving a long-visited discussion here... (e.g. [1] )
Yes, I suspect you are :-)
But as far as sameas.org itself is concerned, the story is a bit more pragmatic.
Note that sameas.org is only the name of a site - think of it as an opaque URI - I called it that because it conveyed something of the nature of the service that people might understand - I could just as easily have called is closeMatch or foobar, but that is not so snappy.
Now the predicate.
It is true that you can get owl:sameas as the predicate returned for some of the result formats; I would be happy to return any predicate and indeed the consumer is at liberty to rewrite the stuff that comes back to any predicate they like before using it.
In fact I am intending to add a way of specifying the predicate in the invocation, but as no-one has yet asked, it is not a high priority.
And actually the early versions of the service did not return RDF at all, and so had no predicate.

I actually view sameas.org as a search service.
As such, I take a very liberal view of equivalence - I have harvested quite a lot of other predicates (dbpedia:redirects, skos:closeMatch and would even do owl:equivalentProperty if there many of them).
The idea is that when you are searching, you want the net cast quite widely.

However, for other sites, such as Freebase and British Library, the interpretation is much less liberal, and is in fact very conservative.
So you can view what comes back as a strong statement of opinion, and it is likely to conform to what you would expect as owl:sameas.

I hope that clarifies things, at least from the sameas.org point of view.

The site name of sameas.org and the owl:sameas predicate are essentially quick ways of telling consumers what it might do for them, and it should not be read into them that there has been any chance in the community's view of how owl:sameAs might be used.

Best
Hugh

> 
> Is it current practice now to use owl:sameAs for mapping ontologies, despite the semantic implications of sameAs? Is a weaker interpretation of sameAs now being taken? 
> 
> My understanding was that alternatives to sameAs should be used unless there is absolutely no doubt that the two Things are in fact exactly the same entity. The vast amount of data in sameas.org suggests to me, though, that common practice is tending towards using (a slightly weaker interpretation of) sameAs ?
> 
> anna
> 
> [1] http://ontologydesignpatterns.org/wiki/Community:Overloading_OWL_sameAs
> 
> On 8 Feb 2012, at 11:13, Hugh Glaser wrote:
> 
>> Hi.
>> 
>> With the addition of Freebase and British Library data (thank you), it is exciting to tell you that http://sameas.org has now topped 100M URIs.
>> 
>> I am always happy to get more data, and can bring up individual sub-stores for people who want them.
>> So please feel free to ask me to add your equivalence information to the 100M!
>> 
>> In case you don't know, there are already some sub-stores that provide results based on subsets of the main store, which are more conservative in their view of URI equivalence.
>> (So they are more statements of a provider's view as opposed to the search provided by the main store.)
>> In particular:
>> http://sameas.org/store/freebase/ - Data from Freebase
>> http://sameas.org/store/britishlibrary/ - Data from the British Library
>> http://sameas.org/store/kelle/ - A specialist collection of data from a number of libraries
>> 
>> And if you want any help using sameas.org, please feel free to ask.
>> 
>> Best
>> Hugh
>> -- 
>> Hugh Glaser,  
>>            Web and Internet Science
>>            Electronics and Computer Science,
>>            University of Southampton,
>>            Southampton SO17 1BJ
>> Work: +44 23 8059 3670, Fax: +44 23 8059 3045
>> Mobile: +44 75 9533 4155 , Home: +44 23 8061 5652
>> http://www.ecs.soton.ac.uk/~hg/
>> 
>> 
> 
> ---
> Anna Jordanous
> Research Associate
> Centre for e-Research
> King's College London
> +44 (0)20 7848 1988
> 

-- 
Hugh Glaser,  
             Web and Internet Science
             Electronics and Computer Science,
             University of Southampton,
             Southampton SO17 1BJ
Work: +44 23 8059 3670, Fax: +44 23 8059 3045
Mobile: +44 75 9533 4155 , Home: +44 23 8061 5652
http://www.ecs.soton.ac.uk/~hg/

Received on Thursday, 9 February 2012 12:21:30 UTC