Re: [DBpedia-discussion] DBpedia URI changes

2018-08-06 Thread Sebastian Hellmann 

Hi Michael,

hm, yes, seems down. We tried to deploy it on a more powerful server at 
http://vmdbpedia.informatik.uni-leipzig.de:5000/same-thing/lookup/?uri=http://dbpedia.org/resource/Brown_Pelican


But this also seems not to deployed correctly.


Any idea when there will be a data dump available with those IDs?
We have some here: 
http://downloads.dbpedia.org/databus/global/persistence-core/


But these are just the prototype, we are trying to oil the engine at the 
moment so things run more smoothly. The Databus Maven Plugin is almost 
finished:


https://github.com/dbpedia/databus-maven-plugin

With this we are aiming to release everything much more frequently and 
have automatic downloads, versioning and RSS feeds.


All the best,

Sebastian


On 06.08.2018 12:13, Michael Brunnbauer wrote:

Hello Sebastian,

On Tue, Jul 31, 2018 at 12:31:13PM +0200, Sebastian Hellmann wrote:

We have created new URIs that are stable. Alex Olieman has already created a
service to map them:

https://e.hum.uva.nl/same-thing/lookup/?uri=http://www.wikidata.org/entity/Q8087

[...]

https://e.hum.uva.nl/same-thing/lookup/?uri=http://dbpedia.org/resource/Brown_pelican
https://e.hum.uva.nl/same-thing/lookup/?uri=http://dbpedia.org/resource/Brown_Pelican

I get a "502 Bad Gateway" error when accessing those links.


- the global ids will be http://global.dbpedia.org/id/${base58endodedBigInt}
we had a conflict with id.dbpedia.org

[...]

As you are updating at the moment, I would suggest that you store the new
ids alongside in your app, which will make further updates much easier.

Any idea when there will be a data dump available with those IDs?

Regards,

Michael Brunnbauer



--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot


___
DBpedia-discussion mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/dbpedia-discussion


--
All the best,
Sebastian Hellmann

Director of Knowledge Integration and Linked Data Technologies (KILT) 
Competence Center

at the Institute for Applied Informatics (InfAI) at Leipzig University
Executive Director of the DBpedia Association
Projects: http://dbpedia.org, http://nlp2rdf.org, 
http://linguistics.okfn.org, https://www.w3.org/community/ld4lt 


Homepage: http://aksw.org/SebastianHellmann
Research Group: http://aksw.org
--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot___
DBpedia-discussion mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/dbpedia-discussion

Re: [DBpedia-discussion] DBpedia URI changes

2018-08-06 Thread Michael Brunnbauer 

Hello Sebastian,

On Tue, Jul 31, 2018 at 12:31:13PM +0200, Sebastian Hellmann wrote:
> We have created new URIs that are stable. Alex Olieman has already created a
> service to map them:
> 
> https://e.hum.uva.nl/same-thing/lookup/?uri=http://www.wikidata.org/entity/Q8087
[...]
> https://e.hum.uva.nl/same-thing/lookup/?uri=http://dbpedia.org/resource/Brown_pelican
> https://e.hum.uva.nl/same-thing/lookup/?uri=http://dbpedia.org/resource/Brown_Pelican

I get a "502 Bad Gateway" error when accessing those links.

> - the global ids will be http://global.dbpedia.org/id/${base58endodedBigInt}
> we had a conflict with id.dbpedia.org
[...]
> As you are updating at the moment, I would suggest that you store the new
> ids alongside in your app, which will make further updates much easier.

Any idea when there will be a data dump available with those IDs?

Regards,

Michael Brunnbauer

-- 
++  Michael Brunnbauer
++  netEstate GmbH
++  Geisenhausener Straße 11a
++  81379 München
++  Tel +49 89 32 19 77 80
++  Fax +49 89 32 19 77 89 
++  E-Mail [email protected]
++  https://www.netestate.de/
++
++  Sitz: München, HRB Nr.142452 (Handelsregister B München)
++  USt-IdNr. DE221033342
++  Geschäftsführer: Michael Brunnbauer, Franz Brunnbauer
++  Prokurist: Dipl. Kfm. (Univ.) Markus Hendel


signature.asc
Description: PGP signature
--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot___
DBpedia-discussion mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/dbpedia-discussion

Re: [DBpedia-discussion] DBpedia URI changes

2018-08-01 Thread Dimitris Kontokostas 
Hi,

Redirects in Wikipedia can mean 3 different things
1. a proper renaming of the title, i.e. capitalization
2. the article is merged with other related articles and the redirect goes
to a subsection of the new article
3. the article is split in multiple, more fine-grained articles (not sure
which new article redirects point to in this case)

in the case of (1), it would be safe to assume something equivalent to
owl:sameAs
In (2) and (3) though not, that would lead to incorrect entity assignments

Taking only one language (i.e. English Wikipedia), it is hard to automate
this process but in the last release there was a kind of different approach
by using wikidata links
Since wikidata links can also link to redirects in a language, a transitive
redirect closure was computed up to a redirect that had a wikidata link and
creating a new closure for the remaining redirects

for example

articleA - redirects to - Article_A - redirects to - Articles_of_A_and_B
articleB - redirects to - Article_B - redirects to - Articles_of_A_and_B

With the previous approach, both articleA and articleB would redirect to
Articles_of_A_and_B
in this case assuming owl:sameAs would not be correct

#New approach
Assuming there exists a wikidata item X1 that links to Article_A and X2
that links to Article_B

only the following redirects are computed
articleA redirects to Article_A
articleB redirects to Article_B

Something like this would probably fit for what you want
If you stick to English, there could be cases where you end up with
entities that no longer exists (articles got merged or split) but would be
very easy to identify them

As far as I know, this algorithm was not deployed in the language specific
version but only on the normalized datasets that are meant for data fusion
(or the DBpedia data bus)
porting this dataset back to English DBpedia should not be a very hard
task, maybe talking to the DBpedia association would be the easiest way to
see available options for a custom dataset export

Best,
Dimitris

On Tue, Jul 31, 2018 at 1:32 PM Sebastian Hellmann <
[email protected]> wrote:

> Hi Michael,
>
> regarding the redirects, it would really be a lot of work to code
> something that distinguishes them into owl:sameAs and rdfs:seeAlso.
>
> That said, I am sure that somebody already did this, see e.g. here:
> https://scholar.google.de/scholar?hl=en&as_sdt=0%2C5&q=wikipedia+redirects&btnG=
> maybe this one: Analyzing and accessing Wikipedia as a lexical semantic
> resource
>
> At the moment we are facing two major challenges, which will hopefully be
> fixed in the next weeks:
>
> 1. it is hard to contribute to DBpedia: we are developing a client to push
> external efforts to DBpedia via the DBpedia Databus, see here:
> https://github.com/dbpedia/data-release-tool-ti/blob/master/README.md
>
> So in the future, anybody who creates such an extension or a let's call it
> a fork of the "DBpedia redirects" dataset can push it again. Ideally, we
> release metadata about the releases and others can scan for new data and
> then run their tools automatically and re-publish results. We recently had
> a call for first people to help us test the release client:
> https://sourceforge.net/p/dbpedia/mailman/message/36375367/
>
> 2. URIs from Wikipedia are not stable in the first place. Not sure about
> the percentage, but let's say 5% of the uris get changed and redirected
> each year
>
> We have created new URIs that are stable. Alex Olieman has already created
> a service to map them:
>
>
> https://e.hum.uva.nl/same-thing/lookup/?uri=http://www.wikidata.org/entity/Q8087
>
> But you can see that the service is also not using the redirects for the
> reasons stated by Nicolas:
>
>
> https://e.hum.uva.nl/same-thing/lookup/?uri=http://dbpedia.org/resource/Brown_pelican
>
>
> https://e.hum.uva.nl/same-thing/lookup/?uri=http://dbpedia.org/resource/Brown_Pelican
>
> Some notes about the service:
>
> - the global ids will be http://global.dbpedia.org/id/${base58endodedBigInt}
> we had a conflict with id.dbpedia.org
>
> - the ids are generated by enumerating all uris af the sources with
> bigints and then merge them into clusters based on owl:sameAs
>
> While this works great per se, we still need a better control over the
> owl:sameAs links that go into the merge. But the effect is that the IDs are
> stable.
>
>
> As you are updating at the moment, I would suggest that you store the new
> ids alongside in your app, which will make further updates much easier.
>
> In general, the data fused in the new releases will be much richer as they
> combine data from all wikipedias and wikidata into one space.
>
> All the best,
>
> Sebastian
>
>
>
>
> On 31.07.2018 09:30, Michael Brunnbauer wrote:
>
>
> Hello Nicolas,
>
> sorry I cannot follow your example. The Terminator6 URI does not seem to
> reference anything on DBpedia:
>
>  http://dbpedia.org/resource/Terminator6
>
> Did it reference something in some earlier version? Not in mine fro

Re: [DBpedia-discussion] DBpedia URI changes

2018-07-31 Thread Sebastian Hellmann 

Hi Michael,

regarding the redirects, it would really be a lot of work to code 
something that distinguishes them into owl:sameAs and rdfs:seeAlso.


That said, I am sure that somebody already did this, see e.g. here: 
https://scholar.google.de/scholar?hl=en&as_sdt=0%2C5&q=wikipedia+redirects&btnG= 
maybe this one: Analyzing and accessing Wikipedia as a lexical semantic 
resource


At the moment we are facing two major challenges, which will hopefully 
be fixed in the next weeks:


1. it is hard to contribute to DBpedia: we are developing a client to 
push external efforts to DBpedia via the DBpedia Databus, see here: 
https://github.com/dbpedia/data-release-tool-ti/blob/master/README.md


So in the future, anybody who creates such an extension or a let's call 
it a fork of the "DBpedia redirects" dataset can push it again. Ideally, 
we release metadata about the releases and others can scan for new data 
and then run their tools automatically and re-publish results. We 
recently had a call for first people to help us test the release client: 
https://sourceforge.net/p/dbpedia/mailman/message/36375367/


2. URIs from Wikipedia are not stable in the first place. Not sure about 
the percentage, but let's say 5% of the uris get changed and redirected 
each year


We have created new URIs that are stable. Alex Olieman has already 
created a service to map them:


https://e.hum.uva.nl/same-thing/lookup/?uri=http://www.wikidata.org/entity/Q8087

But you can see that the service is also not using the redirects for the 
reasons stated by Nicolas:


https://e.hum.uva.nl/same-thing/lookup/?uri=http://dbpedia.org/resource/Brown_pelican

https://e.hum.uva.nl/same-thing/lookup/?uri=http://dbpedia.org/resource/Brown_Pelican

Some notes about the service:

- the global ids will be 
http://global.dbpedia.org/id/${base58endodedBigInt} we had a conflict 
with id.dbpedia.org


- the ids are generated by enumerating all uris af the sources with 
bigints and then merge them into clusters based on owl:sameAs


While this works great per se, we still need a better control over the 
owl:sameAs links that go into the merge. But the effect is that the IDs 
are stable.



As you are updating at the moment, I would suggest that you store the 
new ids alongside in your app, which will make further updates much easier.


In general, the data fused in the new releases will be much richer as 
they combine data from all wikipedias and wikidata into one space.


All the best,

Sebastian




On 31.07.2018 09:30, Michael Brunnbauer wrote:


Hello Nicolas,

sorry I cannot follow your example. The Terminator6 URI does not seem to
reference anything on DBpedia:

  http://dbpedia.org/resource/Terminator6

Did it reference something in some earlier version? Not in mine from 2012.

Regards,

Michael Brunnbauer

On Fri, Jul 27, 2018 at 03:19:08PM -0700, Nicolas Torzec wrote:

I'll let the DBpedia folks reply but identity management is a hard problem.

Example:
- On one hand Brown_Pelican

redirects to Brown_pelican 
and they happen to reference the same entity.
- On the other hand, Terminator 6

redirects to the Terminator franchise
 and yet they
reference different entities.
- So an owl:sameAs  may be
appropriate in the first case but not the second one.

Cheers.
-N.


On Fri, Jul 27, 2018 at 10:20 AM Michael Brunnbauer 
wrote:


hi all,

we are planning to upgrade the DBpedia version used for our app from 3.8
(2012) to the current version.

It seems that when a Wikipedia entry changes its URL, so does the
corresponding Dbpedia URI. This is of course a no go and I don't know why
the team decided not to keep the old URI in that case but I'm sure there
were good reasons.

So the currently used DBpedia version 3.8 has both

  http://dbpedia.org/resource/Brown_Pelican

  http://dbpedia.org/resource/Brown_pelican

with no triples connecting each other. The users of the app I am
maintaining have used the Brown_Pelican URI which seems to be the older one.

The current DBpedia has made Brown_pelican canonical and has the following
to say about the old URI:


http://dbpedia.org/sparql?default-graph-uri=http%3A%2F%2Fdbpedia.org&query=select+%3Fp+%3Fo+where+%7B%0D%0A%3Chttp%3A%2F%2Fdbpedia.org%2Fresource%2FBrown_Pelican%3E+%3Fp+%3Fo%0D%0A%7D&format=text%2Fhtml&CXML_redir_for_subjs=121&CXML_redir_for_hrefs=&timeout=3&debug=on&run=+Run+Query+

There is no owl:sameAs link but a dbo:wikiPageRedirects from the old to
the new URI.

Why is there no owl:sameAs triple generated for every
dbo:wikiPageRedirects?

What possible problems could arise if our app treats dbo:wikiPageRedirects
like a owl:sameAs assertion?

Is it possible that URIs changed without a corresponding
dbo:wikiPageRedir

Re: [DBpedia-discussion] DBpedia URI changes

2018-07-31 Thread Michael Brunnbauer 

Hello Nicolas,

sorry I cannot follow your example. The Terminator6 URI does not seem to
reference anything on DBpedia:

 http://dbpedia.org/resource/Terminator6

Did it reference something in some earlier version? Not in mine from 2012.

Regards,

Michael Brunnbauer

On Fri, Jul 27, 2018 at 03:19:08PM -0700, Nicolas Torzec wrote:
> I'll let the DBpedia folks reply but identity management is a hard problem.
> 
> Example:
> - On one hand Brown_Pelican
> 
> redirects to Brown_pelican 
> and they happen to reference the same entity.
> - On the other hand, Terminator 6
> 
> redirects to the Terminator franchise
>  and yet they
> reference different entities.
> - So an owl:sameAs  may be
> appropriate in the first case but not the second one.
> 
> Cheers.
> -N.
> 
> 
> On Fri, Jul 27, 2018 at 10:20 AM Michael Brunnbauer 
> wrote:
> 
> >
> > hi all,
> >
> > we are planning to upgrade the DBpedia version used for our app from 3.8
> > (2012) to the current version.
> >
> > It seems that when a Wikipedia entry changes its URL, so does the
> > corresponding Dbpedia URI. This is of course a no go and I don't know why
> > the team decided not to keep the old URI in that case but I'm sure there
> > were good reasons.
> >
> > So the currently used DBpedia version 3.8 has both
> >
> >  http://dbpedia.org/resource/Brown_Pelican
> >
> >  http://dbpedia.org/resource/Brown_pelican
> >
> > with no triples connecting each other. The users of the app I am
> > maintaining have used the Brown_Pelican URI which seems to be the older one.
> >
> > The current DBpedia has made Brown_pelican canonical and has the following
> > to say about the old URI:
> >
> >
> > http://dbpedia.org/sparql?default-graph-uri=http%3A%2F%2Fdbpedia.org&query=select+%3Fp+%3Fo+where+%7B%0D%0A%3Chttp%3A%2F%2Fdbpedia.org%2Fresource%2FBrown_Pelican%3E+%3Fp+%3Fo%0D%0A%7D&format=text%2Fhtml&CXML_redir_for_subjs=121&CXML_redir_for_hrefs=&timeout=3&debug=on&run=+Run+Query+
> >
> > There is no owl:sameAs link but a dbo:wikiPageRedirects from the old to
> > the new URI.
> >
> > Why is there no owl:sameAs triple generated for every
> > dbo:wikiPageRedirects?
> >
> > What possible problems could arise if our app treats dbo:wikiPageRedirects
> > like a owl:sameAs assertion?
> >
> > Is it possible that URIs changed without a corresponding
> > dbo:wikiPageRedirects triple documenting the change?
> >
> > Regards,
> >
> > Michael Brunnbauer
> >
> > --
> > ++  Michael Brunnbauer
> > ++  netEstate GmbH
> > ++  Geisenhausener Straße 11a
> > ++  81379 München
> > ++  Tel +49 89 32 19 77 80
> > ++  Fax +49 89 32 19 77 89
> > ++  E-Mail [email protected]
> > ++  https://www.netestate.de/
> > ++
> > ++  Sitz: München, HRB Nr.142452 (Handelsregister B München)
> > ++  USt-IdNr. DE221033342
> > ++  Geschäftsführer: Michael Brunnbauer, Franz Brunnbauer
> > ++  Prokurist: Dipl. Kfm. (Univ.) Markus Hendel
> >
> > --
> > Check out the vibrant tech community on one of the world's most
> > engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> > ___
> > DBpedia-discussion mailing list
> > [email protected]
> > https://lists.sourceforge.net/lists/listinfo/dbpedia-discussion
> >

-- 
++  Michael Brunnbauer
++  netEstate GmbH
++  Geisenhausener Straße 11a
++  81379 München
++  Tel +49 89 32 19 77 80
++  Fax +49 89 32 19 77 89 
++  E-Mail [email protected]
++  https://www.netestate.de/
++
++  Sitz: München, HRB Nr.142452 (Handelsregister B München)
++  USt-IdNr. DE221033342
++  Geschäftsführer: Michael Brunnbauer, Franz Brunnbauer
++  Prokurist: Dipl. Kfm. (Univ.) Markus Hendel


signature.asc
Description: PGP signature
--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot___
DBpedia-discussion mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/dbpedia-discussion