W3C home > Mailing lists > Public > public-lod@w3.org > January 2013

Re: [Dbpedia-discussion] [ANN] Add your links to DBpedia workflow version 0.1 (this is also an RFC)

From: Hugh Glaser <hg@ecs.soton.ac.uk>
Date: Tue, 29 Jan 2013 16:12:34 +0000
To: Sebastian Hellmann <hellmann@informatik.uni-leipzig.de>
CC: Sarven Capadisli <info@csarven.ca>, Søren Roug <Soren.Roug@eea.europa.eu>, public-lod <public-lod@w3.org>
Message-ID: <387E72E216DF1247A2F8ED4819C93BA71E516A96@UOS-MSG00041-SI.soton.ac.uk>

On 24 Jan 2013, at 11:30, Sebastian Hellmann <hellmann@informatik.uni-leipzig.de> wrote:

> There has been some work before in partitioning Wikipedia articles into Instances and Concepts.
> However this is a highly disputed topic. I would go with these being skos:Concepts :
> http://en.wikipedia.org/wiki/Jaguar
> http://en.wikipedia.org/wiki/Country
> http://en.wikipedia.org/wiki/Leg
> 
> and these being instances:
> http://en.wikipedia.org/wiki/Barack_obama
> http://en.wikipedia.org/wiki/Empire_state_building
> 
> with these being problematic:
> http://en.wikipedia.org/wiki/Luke_Skywalker
> http://en.wikipedia.org/wiki/Roman_Empire
> 
> I would not like to discuss this in detail as it might lead to nothing. Please present a use case. Just to be able to use skos:*Match is not sufficient.  Maybe, we can make "usefullness" the main criteria.  Besides, this might be also something, that you want to tackle in the Wikipedia Category system, directly ;)
> 
> So currently, we have owl:sameAs, skos:*Match and umbel:isLike. Any more?
I think there is one in some ontology about bio (is it somewhere around genes?), but I can't for the life of me remember.

Also, what about owl:equivalentClass ?
and even owl:equivalentProperty ?
If you are having skos:closeMatch, then these seem close enough.

I have a feeling I should probably duck now.
> 
> I wouldn't disallow any of these.
> 
> Sebastian
> 
> 
> 
> Am 18.01.2013 13:14, schrieb Sarven Capadisli:
>> Hi Søren,
>> 
>> Yes, that's all valid. The intention was to shift the focus from using owl:sameAs for most things to skos:*Match where they'd be more appropriate. Ideally we'll have all sorts of SKOS Collections, ConceptSchemes, CodeLists for the Concepts which they can be better targeted. We wouldn't just stop at marking dbo:Countrys also with skos:Concepts - it is the first step ;)
>> 
>> -Sarven
>> 
>> On 01/17/2013 09:07 PM, Søren Roug wrote:
>>> I'm not sure I get the usefulness of skos:Concept in DBPedia. To get
>>> the list of countries you can do
>>> 
>>> SELECT ?s WHERE {?s a dbo:Country}
>>> 
>>> If you do a SELECT ?s WHERE {?s a skos:Concept} you would get
>>> whatever someone decided is a skos:Concept. In principle you could
>>> get all entities in the database. To make it useful you'd have to
>>> make skos:Collections of all the classes (i.e. dbo:Country) and then
>>> use skos:member to connect it all.
>>> 
>>> The skos:*Match predicates don't have domain/range properties, so
>>> they can be used outside SKOS vocabularies without side-effects.
>>> 
>>> -- Best regards, Søren Roug ________________________________________
>>> Fra: Sarven Capadisli [info@csarven.ca] Sendt: 17. januar 2013 16:55
>>> Til: Søren Roug; 'Sebastian Hellmann' Cc: Dbpedia-developers; DBpedia
>>> discussion; public-lod Emne: Re: [Dbpedia-discussion] [ANN] Add your
>>> links to DBpedia workflow   version 0.1 (this is also an RFC)
>>> 
>>> I also think this quite important.
>>> 
>>> What I would really like to see is certain conceptual things in
>>> dbpedia is typed with skos:Concept. That gives me some assurance and
>>> encourages me (and probably others) to use skos:*Match where
>>> appropriate instead of slapping owl:sameAs.
>>> 
>>> For instance, resources that are of type dbo:Country, can we have
>>> skos:Concept for these for starters? :)
>>> 
>>> -Sarven
>>> 
>>> 
>>> On 01/17/2013 01:22 PM, Søren Roug wrote:
>>>> Hi,
>>>> 
>>>> I see that you have also added EUNIS links for species. I can
>>>> provide you with a much better quality, and I have forked the repo.
>>>> But I am concerned about the use of owl:sameAs as the predicate.
>>>> The community doesn't have consensus on the species names and we
>>>> can't even agree on many species - e.g. is the goat in the Pyrenees
>>>> the same species as the one in southern Spain or a subspecies? I
>>>> don't want some OWL-aware query tool to be utterly confused.
>>>> 
>>>> What I'm asking is: Can I use skos:exactMatch and skos:closeMatch
>>>> instead of owl:sameAs?
>>>> 
>>>> -- Sincerely yours / Med venlig hilsen, Søren Roug
>>>> <soren.roug@eea.europa.eu> European Environment Agency, Kongens
>>>> Nytorv 6, DK-1050 Copenhagen K Tel: +45 2368 3660 Jabber:
>>>> roug@jabber.eea.europa.eu This email was delivered using 100%
>>>> recycled electrons. Please try to keep it that way.
>>>> 
>>>> 
>>>> || -----Original Message----- || From: Sebastian Hellmann
>>>> [mailto:hellmann@informatik.uni-leipzig.de] || Sent: 17 January
>>>> 2013 10:18 || To: Dbpedia-developers; DBpedia discussion;
>>>> public-lod || Subject: Re: [Dbpedia-discussion] [ANN] Add your
>>>> links to DBpedia workflow || version 0.1 (this is also an RFC) ||
>>>> || Nice, we received links to 2 data sets yesterday, already,
>>>> thanks to || Sarven: http://270a.info || It is a good start and we
>>>> are really hoping for more. Note that most of || the links in
>>>> http://richard.cyganiak.de/2007/10/lod/imagemap.html only || point
>>>> to DBpedia and not outwards. || || In case you are not a Git
>>>> expert, GitHub allows you to upload links with || their GUI: || 1.
>>>> create an account || 2. fork this repo
>>>> https://github.com/dbpedia/dbpedia-links || 3. click on the "create
>>>> a file button" and upload your links || 4. send a "pull request" ||
>>>> || We might simplify the way to contribute later. Right now, Git
>>>> seems to || be the easiest way. Any ideas are welcome. || all the
>>>> best, || Sebastian || || || || Am 16.01.2013 15:06, schrieb
>>>> Sebastian Hellmann: || > Dear all, || > || > we thought that it
>>>> might be a nice idea to simplify the workflow for || > creating
>>>> outgoing links from DBpedia to your data sets. This is why we || >
>>>> created the following GitHub repository: || >
>>>> https://github.com/dbpedia/dbpedia-links ||
>>>>> || > Please feel free to add new files and change the links and
>>>> then send us || > a *pull request*.   This message is an
>>>> announcement as well as a request || > for comments. || > || > Here
>>>> is a (non-exhaustive) list of open issue: || > || > - it is yet
>>>> unclear, when the links will be loaded into || >
>>>> http://dbpedia.org/sparql (maybe with version 3.9? ) || > - we plan
>>>> weekly updates to http://live.dbpedia.org || > - yago, freebase and
>>>> flickrwrappr have been excluded due to their size ( || > > 0.5GB
>>>> ) || > - there will be some quality control; not everybody will be
>>>> able to || > include any links he wants to include. We are open to
>>>> ideas how to || > manage this. Consider "pull requests" as
>>>> "application for inclusion" || > - folder/file structure is still
>>>> very simple, we will adapt upon uptake || > || > All the best, || >
>>>> Sebastian || > || > || > || || || -- || Dipl. Inf. Sebastian
>>>> Hellmann || Department of Computer Science, University of Leipzig
>>>> || Projects: http://nlp2rdf.org , http://dbpedia.org || Homepage:
>>>> http://bis.informatik.uni-leipzig.de/SebastianHellmann || Research
>>>> Group: http://aksw.org
>>>> 
>>>> 
>>>> 
>>>> 
>>> 
>>> 
>> 
>> 
>> 
> 
> 
> -- 
> Dipl. Inf. Sebastian Hellmann
> Department of Computer Science, University of Leipzig
> Projects: http://nlp2rdf.org , http://dbpedia.org
> Homepage: http://bis.informatik.uni-leipzig.de/SebastianHellmann
> Research Group: http://aksw.org
> 
Received on Tuesday, 29 January 2013 16:14:18 UTC

This archive was generated by hypermail 2.3.1 : Sunday, 31 March 2013 14:24:45 UTC