Re: [Dbpedia-discussion] A quick analysis of the classes in the DBpedia ontology
The German chapter is working on ways to automatically import axioms in
the mappings wiki, so of course this is an option too.
We can more or less do this since we are already inserting labels into the
ontology in batch mode [1]. It would be quite helpful for us if someone
interested in editing the ontology programmatically, would produce a list
of changes (class name and changed properties, in any kind of machine
readable format you want, even Sparql would be nice). We could then better
experiment and introduce the changes directly into the mappings wiki.
Changing the ontology in the triplestore would create a synchronization
problem, by the next extraction those changes would need to be recomputed
and reintroduced.
Cheers,
Alexandru
[1] https://github.com/ag-csw/missingBot
On Tue, Apr 15, 2014 at 8:06 AM, Dimitris Kontokostas jimk...@gmail.comwrote:
Kingsley's approach is one way to go but I think we should focus on fixing
the ontology in the source, which is the mappings wiki.
The German chapter is working on ways to automatically import axioms in
the mappings wiki, so of course this is an option too.
I think you already pointed out most of your suggested changes so we can
start working on these. [1]
Cheers,
Dimitris
[1]
http://www.mail-archive.com/dbpedia-discussion@lists.sourceforge.net/msg05923.html
On Mon, Apr 14, 2014 at 9:46 PM, Kingsley Idehen
kide...@openlinksw.comwrote:
On 4/14/14 12:07 PM, Patel-Schneider, Peter wrote:
Aaah, sure I can use SPARQL 1.1 to massage triple stores, including
triple stores that use IRIs from the DBpedia ontology. In this way, I
could modify the results, perhaps to make them look like certain stuff had
been removed from the DBpedia ontology, although this process can result in
systematic errors if the triples include inferred triples.
My question was, however, what SPARQL 1.1 has to do with changing the
DBpedia ontology itself.
Answer:
It enables you import the data in question, conditionally (via SPARQL
query pattern solution), en route to crafting a new ontology or tweaking
the existing ontology.
You are ultimately going to be doing at least one of the following
(locally):
1. adding new RDF statements
2. deleting existing RDF statements
3. updating existing RDF statements (via conditional INSERT and DELETE).
Of course, you can forget SPARQL and just import the lot and editing by
hand etc..
Fundamentally, you can fix DBpedia's ontology by contributing your fixes
in the form of RDF statements for consideration by the maintainers. If
rejected (for whatever reasons) you can still publish your RDF statements
via an RDF document to some location under your control on the Web.
Your revised ontology is your set of context lenses into the DBepdia
dataset.
If you don't want to craft the new ontology or fixes to the existing
ontology, using the methods suggested, how else can you expect this to
happen?
Kingsley
peter
On Apr 14, 2014, at 5:42 AM, Kingsley Idehen kide...@openlinksw.com
wrote:
On 4/14/14 7:02 AM, Kingsley Idehen wrote:
How can I use SPARQL 1.1 to change the DBpedia ontology?
You can use SPARQL 1.1 (from your SPARQL 1.1 compliant application) to
generate triples is a named graph local to your application, based
solutions returned to you from the public SPARQL endpoint.
Important typo fix. I meant to say:
You can use SPARQL 1.1 (from your SPARQL 1.1 compliant application) to
generate triples *in* a named graph local to your application, based *on*
solutions returned to *your app* from the public SPARQL endpoint.
I do this all the time when editing definitions (for classes and
properties) oriented triples in ontologies.
Typical example, where I am adding missing rdfs:isDefinedBy,
voca:defines, and wdrs:isdescribedby relations to an existing shared
ontology.
Note: I am using Virtuoso as my SPARQL 1.1 compliant application (so I
have the ability to grab data from any SPARQL endpoint and then use SPARQL
1.1 for local named graph scoped INSERT and DELETE operations):
## The Identity of Resources on the Web ontology (IRW).
# Ontology URI: http://www.ontologydesignpatterns.org/ont/web/irw.owl#
# Ontology Document URL: http://www.ontologydesignpatterns.org/
ont/web/irw.owl
## I use SPARQL 1.1 LOAD to get the data into Virtuoso
LOAD http://www.ontologydesignpatterns.org/ont/web/irw.owl ;
WITH GRAPH http://www.ontologydesignpatterns.org/ont/web/irw.owl
INSERT {
?s http://www.w3.org/2000/01/rdf-schema#isDefinedBy
http://www.ontologydesignpatterns.org/ont/web/irw.owl# .
http://www.ontologydesignpatterns.org/ont/web/irw.owl#
http://open.vocab.org/terms/defines ?s.
http://www.ontologydesignpatterns.org/ont/web/irw.owl# a
owl:Ontology .
?s http://www.w3.org/2007/05/powder-s#describedby
http://www.ontologydesignpatterns.org/ont/web/irw.owl .
http://www.ontologydesignpatterns.org/ont/web/irw.owl
foaf:primaryTopic ?s .
}
Re: [Dbpedia-discussion] A quick analysis of the classes in the DBpedia ontology
On 4/15/14 1:06 AM, Dimitris Kontokostas wrote: Kingsley's approach is one way to go but I think we should focus on fixing the ontology in the source, which is the mappings wiki. The German chapter is working on ways to automatically import axioms in the mappings wiki, so of course this is an option too. I think you already pointed out most of your suggested changes so we can start working on these. [1] Cheers, Dimitris [1] http://www.mail-archive.com/dbpedia-discussion@lists.sourceforge.net/msg05923.html We should use this episode to make a clearer guidelines for evolving DBpedia. Options: 1. In the source -- via mapping Wiki 2. Out side the source -- via RDF documents derived from the DBpedia data in question 3. A bit of both -- in situations where the changes have a deeper philosophical bent. What I glean from this episode is that crowd-sourcing options for evolving the DBpedia ontology are not as clear as we generally assume, which ultimately leads to misconceptions etc.. -- Regards, Kingsley Idehen Founder CEO OpenLink Software Company Web: http://www.openlinksw.com Personal Weblog: http://www.openlinksw.com/blog/~kidehen Twitter Profile: https://twitter.com/kidehen Google+ Profile: https://plus.google.com/+KingsleyIdehen/about LinkedIn Profile: http://www.linkedin.com/in/kidehen smime.p7s Description: S/MIME Cryptographic Signature -- Learn Graph Databases - Download FREE O'Reilly Book Graph Databases is the definitive new guide to graph databases and their applications. Written by three acclaimed leaders in the field, this first edition is now available. Download your free book today! http://p.sf.net/sfu/NeoTech___ Dbpedia-discussion mailing list Dbpedia-discussion@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/dbpedia-discussion
Re: [Dbpedia-discussion] A quick analysis of the classes in the DBpedia ontology
On 4/13/14 8:32 PM, Patel-Schneider, Peter wrote: On Apr 11, 2014, at 11:58 AM, Kingsley Idehen kide...@openlinksw.com wrote: On 4/11/14 2:12 PM, Patel-Schneider, Peter wrote: This proposal illustrates one of the major problems with the DBpedia ontology - triples check in but they never check out. Sorta, because of the misconception that SPARQL is steal Read-Only. I spend a good chunk of my day writing SPARQL 1.1 using INSERT, DELETE, and UPDATE (via INSERT and DELETE combos) to massage data, across many data spaces. How can I use SPARQL 1.1 to change the DBpedia ontology? You can use SPARQL 1.1 (from your SPARQL 1.1 compliant application) to generate triples is a named graph local to your application, based solutions returned to you from the public SPARQL endpoint. You can use LOAD, INSERT etc.. to produce your local RDF statements expressing whatever you have in mind. Once done, you can publish an RDF document for incorporation back into the DBpedia project etc.. Many of the problems with the DBpedia ontology require changes. Change is good. And it will work. Remember, DBpedia deploys Linked Data using a Quad Store, so the re-write rules and SPARQL queries used to perform the name-address indirection are extremely flexible. For instance, you can actually set the named graph URI scope for these SPARQL queries explicitly or via our NOT FROM NAMED GRAPH extension re., negation etc.. You can have a list of Named Graphs or excluded Named Graphs when generating the description of a DBpedia Entity URI's referent. I'm not sure how any of this can be used to effect changes in the DBpedia ontology. Until you make a local copy, as I described above, you will believe the statement to be true. All you are doing (ultimately) is make changes in a document and then sending them over for incorporation. In the very worst case, you RDF document content will still be part of the LOD Cloud as long as you publish it on the Web. The first step is making an RDF document with the alternative view that you seek. Other problems have to do with the expressivity of the ontology. Again, its just triples to which SPARQL 1.1 patterns can be applied. In short, this is the way to truly appreciate the power of SPARQL and Linked Data. Again, I'm not sure what play SPARQL 1.1 has with respect to the expressive power of the ontology. SPARQL 1.1 let's you make new RDF statements from existing RDF statements. The DBpedia ontology is a collection of RDF statements. I don't think that changes here can be effected just by adding and removing bits of the ontology. You can variants of the Ontology, an alternative Ontology, it doesn't matter, the Linked Data deployment will be unaffected. Well, sure, none of this will affect most Linked Data uses, but that's not what I'm interested in. I'm interested in using the DBpedia ontology to organize information. Yes, and information lives in documents, right? Thus, you simply grab the relevant data from DBpedia, massage it (using SPARQL 1.1 or other means) and then you have a new document (comprised of new or revised RDF statements). I could, of course, simply use a different ontology, but my hope here is that use of the DBpedia ontology in products will result in a better ontology, and that that can be shared. Yes, why is contributing tweaks to the DBpedia ontology using an RDF document produced by you not an option here? Do that, and everything else falls into place. Other problems have to do with the philosophy of the ontology. The philosophy of the Ontology cannot change, that's a world view of the ontology creators. That doesn't stop another ontology existing as an alternative set of context lenses into the same data. It appears that the current ontology does not match the stated philosophy of the ontology. One or the other should change, and probably both. Can't you reflect that in an RDF document submitted to the project? I encourage you to make your changes, or make a new ontology, whichever path you take, the end product will be useful and a showcase for perspectives sometimes overlooked due to blurred and blurry perspectives :-) I would love to make changes. There actually is a modified version of the ontology that is in use. Is this ontology represented in an RDF document that's accessible via an HTTP URL, at this point in time? If it exists, then we are nearly there. Kingsley peter We can do this, its the next stage in the natural evolution of DBpedia and the broader Linked Open Data Cloud. Note: there is zero speculation in my response. I've already done (and continue to do) a lot of this (hands on fashion) over the years, following LOD cloud initial bootstrap. BTW -- This mail was easy to write as I've just completed (literally minutes ago) a marathon session on SPARQL 1.1 which amounted to an ACL (data access policy and access control list) debugging session. Kingsley.
Re: [Dbpedia-discussion] A quick analysis of the classes in the DBpedia ontology
On 4/14/14 7:02 AM, Kingsley Idehen wrote:
How can I use SPARQL 1.1 to change the DBpedia ontology?
You can use SPARQL 1.1 (from your SPARQL 1.1 compliant application) to
generate triples is a named graph local to your application, based
solutions returned to you from the public SPARQL endpoint.
Important typo fix. I meant to say:
You can use SPARQL 1.1 (from your SPARQL 1.1 compliant application) to
generate triples *in* a named graph local to your application, based
*on* solutions returned to *your app* from the public SPARQL endpoint.
I do this all the time when editing definitions (for classes and
properties) oriented triples in ontologies.
Typical example, where I am adding missing rdfs:isDefinedBy,
voca:defines, and wdrs:isdescribedby relations to an existing shared
ontology.
Note: I am using Virtuoso as my SPARQL 1.1 compliant application (so I
have the ability to grab data from any SPARQL endpoint and then use
SPARQL 1.1 for local named graph scoped INSERT and DELETE operations):
## The Identity of Resources on the Web ontology (IRW).
# Ontology URI: http://www.ontologydesignpatterns.org/ont/web/irw.owl#
# Ontology Document URL:
http://www.ontologydesignpatterns.org/ont/web/irw.owl
## I use SPARQL 1.1 LOAD to get the data into Virtuoso
LOAD http://www.ontologydesignpatterns.org/ont/web/irw.owl ;
WITH GRAPH http://www.ontologydesignpatterns.org/ont/web/irw.owl
INSERT {
?s http://www.w3.org/2000/01/rdf-schema#isDefinedBy
http://www.ontologydesignpatterns.org/ont/web/irw.owl# .
http://www.ontologydesignpatterns.org/ont/web/irw.owl#
http://open.vocab.org/terms/defines ?s.
http://www.ontologydesignpatterns.org/ont/web/irw.owl# a owl:Ontology .
?s http://www.w3.org/2007/05/powder-s#describedby
http://www.ontologydesignpatterns.org/ont/web/irw.owl .
http://www.ontologydesignpatterns.org/ont/web/irw.owl
foaf:primaryTopic ?s .
}
WHERE {
{?s http://www.w3.org/2000/01/rdf-schema#subClassOf ?o}
UNION
{?s http://www.w3.org/2000/01/rdf-schema#subPropertyOf ?o}
UNION
{?s http://www.w3.org/2002/07/owl#equivalentClass ?o}
UNION
{?s http://www.w3.org/2002/07/owl#equivalentProperty ?o}
UNION
{?s a ?o}
UNION {
?s http://www.w3.org/2000/01/rdf-schema#domain ?o}
UNION
{?s http://www.w3.org/2000/01/rdf-schema#range ?o}
}
## COSMO Ontology
## Ontology URI: http://micra.com/COSMO/COSMO.owl#
## Ontology Document URL: http://micra.com/COSMO/COSMO.owl
## Using SPARQL 1.1 LOAD to get data into Virtuoso
LOAD http://micra.com/COSMO/COSMO.owl ;
WITH GRAPH http://micra.com/COSMO/COSMO.owl
INSERT {
?s http://www.w3.org/2000/01/rdf-schema#isDefinedBy
http://micra.com/COSMO/COSMO.owl# .
http://micra.com/COSMO/COSMO.owl#
http://open.vocab.org/terms/defines ?s.
http://micra.com/COSMO/COSMO.owl# a owl:Ontology .
?s http://www.w3.org/2007/05/powder-s#describedby
http://micra.com/COSMO/COSMO.owl .
http://micra.com/COSMO/COSMO.owl foaf:primaryTopic ?s .
}
WHERE {
{?s http://www.w3.org/2000/01/rdf-schema#subClassOf ?o}
UNION
{?s http://www.w3.org/2000/01/rdf-schema#subPropertyOf ?o}
UNION
{?s http://www.w3.org/2002/07/owl#equivalentClass ?o}
UNION
{?s http://www.w3.org/2002/07/owl#equivalentProperty ?o}
UNION
{?s a ?o}
UNION {
?s http://www.w3.org/2000/01/rdf-schema#domain ?o}
UNION
{?s http://www.w3.org/2000/01/rdf-schema#range ?o}
}
--
Regards,
Kingsley Idehen
Founder CEO
OpenLink Software
Company Web: http://www.openlinksw.com
Personal Weblog: http://www.openlinksw.com/blog/~kidehen
Twitter Profile: https://twitter.com/kidehen
Google+ Profile: https://plus.google.com/+KingsleyIdehen/about
LinkedIn Profile: http://www.linkedin.com/in/kidehen
smime.p7s
Description: S/MIME Cryptographic Signature
--
Learn Graph Databases - Download FREE O'Reilly Book
Graph Databases is the definitive new guide to graph databases and their
applications. Written by three acclaimed leaders in the field,
this first edition is now available. Download your free book today!
http://p.sf.net/sfu/NeoTech___
Dbpedia-discussion mailing list
Dbpedia-discussion@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/dbpedia-discussion
Re: [Dbpedia-discussion] A quick analysis of the classes in the DBpedia ontology
Aaah, sure I can use SPARQL 1.1 to massage triple stores, including triple
stores that use IRIs from the DBpedia ontology. In this way, I could modify
the results, perhaps to make them look like certain stuff had been removed from
the DBpedia ontology, although this process can result in systematic errors if
the triples include inferred triples.
My question was, however, what SPARQL 1.1 has to do with changing the DBpedia
ontology itself.
peter
On Apr 14, 2014, at 5:42 AM, Kingsley Idehen kide...@openlinksw.com wrote:
On 4/14/14 7:02 AM, Kingsley Idehen wrote:
How can I use SPARQL 1.1 to change the DBpedia ontology?
You can use SPARQL 1.1 (from your SPARQL 1.1 compliant application) to
generate triples is a named graph local to your application, based solutions
returned to you from the public SPARQL endpoint.
Important typo fix. I meant to say:
You can use SPARQL 1.1 (from your SPARQL 1.1 compliant application) to
generate triples *in* a named graph local to your application, based *on*
solutions returned to *your app* from the public SPARQL endpoint.
I do this all the time when editing definitions (for classes and properties)
oriented triples in ontologies.
Typical example, where I am adding missing rdfs:isDefinedBy, voca:defines,
and wdrs:isdescribedby relations to an existing shared ontology.
Note: I am using Virtuoso as my SPARQL 1.1 compliant application (so I have
the ability to grab data from any SPARQL endpoint and then use SPARQL 1.1 for
local named graph scoped INSERT and DELETE operations):
## The Identity of Resources on the Web ontology (IRW).
# Ontology URI: http://www.ontologydesignpatterns.org/ont/web/irw.owl#
# Ontology Document URL:
http://www.ontologydesignpatterns.org/ont/web/irw.owl
## I use SPARQL 1.1 LOAD to get the data into Virtuoso
LOAD http://www.ontologydesignpatterns.org/ont/web/irw.owl ;
WITH GRAPH http://www.ontologydesignpatterns.org/ont/web/irw.owl
INSERT {
?s http://www.w3.org/2000/01/rdf-schema#isDefinedBy
http://www.ontologydesignpatterns.org/ont/web/irw.owl# .
http://www.ontologydesignpatterns.org/ont/web/irw.owl#
http://open.vocab.org/terms/defines ?s.
http://www.ontologydesignpatterns.org/ont/web/irw.owl# a owl:Ontology .
?s http://www.w3.org/2007/05/powder-s#describedby
http://www.ontologydesignpatterns.org/ont/web/irw.owl .
http://www.ontologydesignpatterns.org/ont/web/irw.owl foaf:primaryTopic ?s .
}
WHERE {
{?s http://www.w3.org/2000/01/rdf-schema#subClassOf ?o}
UNION
{?s http://www.w3.org/2000/01/rdf-schema#subPropertyOf ?o}
UNION
{?s http://www.w3.org/2002/07/owl#equivalentClass ?o}
UNION
{?s http://www.w3.org/2002/07/owl#equivalentProperty ?o}
UNION
{?s a ?o}
UNION {
?s http://www.w3.org/2000/01/rdf-schema#domain ?o}
UNION
{?s http://www.w3.org/2000/01/rdf-schema#range ?o}
}
## COSMO Ontology
## Ontology URI: http://micra.com/COSMO/COSMO.owl#
## Ontology Document URL: http://micra.com/COSMO/COSMO.owl
## Using SPARQL 1.1 LOAD to get data into Virtuoso
LOAD http://micra.com/COSMO/COSMO.owl ;
WITH GRAPH http://micra.com/COSMO/COSMO.owl
INSERT {
?s http://www.w3.org/2000/01/rdf-schema#isDefinedBy
http://micra.com/COSMO/COSMO.owl# .
http://micra.com/COSMO/COSMO.owl#
http://open.vocab.org/terms/defines ?s.
http://micra.com/COSMO/COSMO.owl# a owl:Ontology .
?s http://www.w3.org/2007/05/powder-s#describedby
http://micra.com/COSMO/COSMO.owl .
http://micra.com/COSMO/COSMO.owl foaf:primaryTopic ?s .
}
WHERE {
{?s http://www.w3.org/2000/01/rdf-schema#subClassOf ?o}
UNION
{?s http://www.w3.org/2000/01/rdf-schema#subPropertyOf ?o}
UNION
{?s http://www.w3.org/2002/07/owl#equivalentClass ?o}
UNION
{?s http://www.w3.org/2002/07/owl#equivalentProperty ?o}
UNION
{?s a ?o}
UNION {
?s http://www.w3.org/2000/01/rdf-schema#domain ?o}
UNION
{?s http://www.w3.org/2000/01/rdf-schema#range ?o}
}
--
Regards,
Kingsley Idehen
Founder CEO
OpenLink Software
Company Web: http://www.openlinksw.com
Personal Weblog: http://www.openlinksw.com/blog/~kidehen
Twitter Profile: https://twitter.com/kidehen
Google+ Profile: https://plus.google.com/+KingsleyIdehen/about
LinkedIn Profile: http://www.linkedin.com/in/kidehen
--
Learn Graph Databases - Download FREE O'Reilly Book
Graph Databases is the definitive new guide to graph databases and their
applications. Written by three acclaimed leaders in the field,
this first edition is now available. Download your free book today!
http://p.sf.net/sfu/NeoTech
___
Dbpedia-discussion mailing list
Re: [Dbpedia-discussion] A quick analysis of the classes in the DBpedia ontology
On 4/14/14 12:07 PM, Patel-Schneider, Peter wrote:
Aaah, sure I can use SPARQL 1.1 to massage triple stores, including triple
stores that use IRIs from the DBpedia ontology. In this way, I could modify
the results, perhaps to make them look like certain stuff had been removed from
the DBpedia ontology, although this process can result in systematic errors if
the triples include inferred triples.
My question was, however, what SPARQL 1.1 has to do with changing the DBpedia
ontology itself.
Answer:
It enables you import the data in question, conditionally (via SPARQL
query pattern solution), en route to crafting a new ontology or tweaking
the existing ontology.
You are ultimately going to be doing at least one of the following
(locally):
1. adding new RDF statements
2. deleting existing RDF statements
3. updating existing RDF statements (via conditional INSERT and DELETE).
Of course, you can forget SPARQL and just import the lot and editing by
hand etc..
Fundamentally, you can fix DBpedia's ontology by contributing your fixes
in the form of RDF statements for consideration by the maintainers. If
rejected (for whatever reasons) you can still publish your RDF
statements via an RDF document to some location under your control on
the Web.
Your revised ontology is your set of context lenses into the DBepdia
dataset.
If you don't want to craft the new ontology or fixes to the existing
ontology, using the methods suggested, how else can you expect this to
happen?
Kingsley
peter
On Apr 14, 2014, at 5:42 AM, Kingsley Idehen kide...@openlinksw.com wrote:
On 4/14/14 7:02 AM, Kingsley Idehen wrote:
How can I use SPARQL 1.1 to change the DBpedia ontology?
You can use SPARQL 1.1 (from your SPARQL 1.1 compliant application) to generate
triples is a named graph local to your application, based solutions returned to
you from the public SPARQL endpoint.
Important typo fix. I meant to say:
You can use SPARQL 1.1 (from your SPARQL 1.1 compliant application) to generate
triples *in* a named graph local to your application, based *on* solutions
returned to *your app* from the public SPARQL endpoint.
I do this all the time when editing definitions (for classes and properties)
oriented triples in ontologies.
Typical example, where I am adding missing rdfs:isDefinedBy, voca:defines, and
wdrs:isdescribedby relations to an existing shared ontology.
Note: I am using Virtuoso as my SPARQL 1.1 compliant application (so I have the
ability to grab data from any SPARQL endpoint and then use SPARQL 1.1 for local
named graph scoped INSERT and DELETE operations):
## The Identity of Resources on the Web ontology (IRW).
# Ontology URI: http://www.ontologydesignpatterns.org/ont/web/irw.owl#
# Ontology Document URL: http://www.ontologydesignpatterns.org/ont/web/irw.owl
## I use SPARQL 1.1 LOAD to get the data into Virtuoso
LOAD http://www.ontologydesignpatterns.org/ont/web/irw.owl ;
WITH GRAPH http://www.ontologydesignpatterns.org/ont/web/irw.owl
INSERT {
?s http://www.w3.org/2000/01/rdf-schema#isDefinedBy
http://www.ontologydesignpatterns.org/ont/web/irw.owl# .
http://www.ontologydesignpatterns.org/ont/web/irw.owl#
http://open.vocab.org/terms/defines ?s.
http://www.ontologydesignpatterns.org/ont/web/irw.owl# a owl:Ontology .
?s http://www.w3.org/2007/05/powder-s#describedby
http://www.ontologydesignpatterns.org/ont/web/irw.owl .
http://www.ontologydesignpatterns.org/ont/web/irw.owl foaf:primaryTopic ?s .
}
WHERE {
{?s http://www.w3.org/2000/01/rdf-schema#subClassOf ?o}
UNION
{?s http://www.w3.org/2000/01/rdf-schema#subPropertyOf ?o}
UNION
{?s http://www.w3.org/2002/07/owl#equivalentClass ?o}
UNION
{?s http://www.w3.org/2002/07/owl#equivalentProperty ?o}
UNION
{?s a ?o}
UNION {
?s http://www.w3.org/2000/01/rdf-schema#domain ?o}
UNION
{?s http://www.w3.org/2000/01/rdf-schema#range ?o}
}
## COSMO Ontology
## Ontology URI: http://micra.com/COSMO/COSMO.owl#
## Ontology Document URL: http://micra.com/COSMO/COSMO.owl
## Using SPARQL 1.1 LOAD to get data into Virtuoso
LOAD http://micra.com/COSMO/COSMO.owl ;
WITH GRAPH http://micra.com/COSMO/COSMO.owl
INSERT {
?s http://www.w3.org/2000/01/rdf-schema#isDefinedBy
http://micra.com/COSMO/COSMO.owl# .
http://micra.com/COSMO/COSMO.owl#
http://open.vocab.org/terms/defines ?s.
http://micra.com/COSMO/COSMO.owl# a owl:Ontology .
?s http://www.w3.org/2007/05/powder-s#describedby
http://micra.com/COSMO/COSMO.owl .
http://micra.com/COSMO/COSMO.owl foaf:primaryTopic ?s .
}
WHERE {
{?s http://www.w3.org/2000/01/rdf-schema#subClassOf ?o}
UNION
{?s http://www.w3.org/2000/01/rdf-schema#subPropertyOf ?o}
UNION
{?s http://www.w3.org/2002/07/owl#equivalentClass ?o}
UNION
{?s
Re: [Dbpedia-discussion] A quick analysis of the classes in the DBpedia ontology
Kingsley's approach is one way to go but I think we should focus on fixing
the ontology in the source, which is the mappings wiki.
The German chapter is working on ways to automatically import axioms in the
mappings wiki, so of course this is an option too.
I think you already pointed out most of your suggested changes so we can
start working on these. [1]
Cheers,
Dimitris
[1]
http://www.mail-archive.com/dbpedia-discussion@lists.sourceforge.net/msg05923.html
On Mon, Apr 14, 2014 at 9:46 PM, Kingsley Idehen kide...@openlinksw.comwrote:
On 4/14/14 12:07 PM, Patel-Schneider, Peter wrote:
Aaah, sure I can use SPARQL 1.1 to massage triple stores, including
triple stores that use IRIs from the DBpedia ontology. In this way, I
could modify the results, perhaps to make them look like certain stuff had
been removed from the DBpedia ontology, although this process can result in
systematic errors if the triples include inferred triples.
My question was, however, what SPARQL 1.1 has to do with changing the
DBpedia ontology itself.
Answer:
It enables you import the data in question, conditionally (via SPARQL
query pattern solution), en route to crafting a new ontology or tweaking
the existing ontology.
You are ultimately going to be doing at least one of the following
(locally):
1. adding new RDF statements
2. deleting existing RDF statements
3. updating existing RDF statements (via conditional INSERT and DELETE).
Of course, you can forget SPARQL and just import the lot and editing by
hand etc..
Fundamentally, you can fix DBpedia's ontology by contributing your fixes
in the form of RDF statements for consideration by the maintainers. If
rejected (for whatever reasons) you can still publish your RDF statements
via an RDF document to some location under your control on the Web.
Your revised ontology is your set of context lenses into the DBepdia
dataset.
If you don't want to craft the new ontology or fixes to the existing
ontology, using the methods suggested, how else can you expect this to
happen?
Kingsley
peter
On Apr 14, 2014, at 5:42 AM, Kingsley Idehen kide...@openlinksw.com
wrote:
On 4/14/14 7:02 AM, Kingsley Idehen wrote:
How can I use SPARQL 1.1 to change the DBpedia ontology?
You can use SPARQL 1.1 (from your SPARQL 1.1 compliant application) to
generate triples is a named graph local to your application, based
solutions returned to you from the public SPARQL endpoint.
Important typo fix. I meant to say:
You can use SPARQL 1.1 (from your SPARQL 1.1 compliant application) to
generate triples *in* a named graph local to your application, based *on*
solutions returned to *your app* from the public SPARQL endpoint.
I do this all the time when editing definitions (for classes and
properties) oriented triples in ontologies.
Typical example, where I am adding missing rdfs:isDefinedBy,
voca:defines, and wdrs:isdescribedby relations to an existing shared
ontology.
Note: I am using Virtuoso as my SPARQL 1.1 compliant application (so I
have the ability to grab data from any SPARQL endpoint and then use SPARQL
1.1 for local named graph scoped INSERT and DELETE operations):
## The Identity of Resources on the Web ontology (IRW).
# Ontology URI: http://www.ontologydesignpatterns.org/ont/web/irw.owl#
# Ontology Document URL: http://www.ontologydesignpatterns.org/
ont/web/irw.owl
## I use SPARQL 1.1 LOAD to get the data into Virtuoso
LOAD http://www.ontologydesignpatterns.org/ont/web/irw.owl ;
WITH GRAPH http://www.ontologydesignpatterns.org/ont/web/irw.owl
INSERT {
?s http://www.w3.org/2000/01/rdf-schema#isDefinedBy
http://www.ontologydesignpatterns.org/ont/web/irw.owl# .
http://www.ontologydesignpatterns.org/ont/web/irw.owl#
http://open.vocab.org/terms/defines ?s.
http://www.ontologydesignpatterns.org/ont/web/irw.owl# a owl:Ontology
.
?s http://www.w3.org/2007/05/powder-s#describedby http://www.
ontologydesignpatterns.org/ont/web/irw.owl .
http://www.ontologydesignpatterns.org/ont/web/irw.owl
foaf:primaryTopic ?s .
}
WHERE {
{?s http://www.w3.org/2000/01/rdf-schema#subClassOf ?o}
UNION
{?s http://www.w3.org/2000/01/rdf-schema#subPropertyOf ?o}
UNION
{?s http://www.w3.org/2002/07/owl#equivalentClass ?o}
UNION
{?s http://www.w3.org/2002/07/owl#equivalentProperty ?o}
UNION
{?s a ?o}
UNION {
?s http://www.w3.org/2000/01/rdf-schema#domain ?o}
UNION
{?s http://www.w3.org/2000/01/rdf-schema#range ?o}
}
## COSMO Ontology
## Ontology URI: http://micra.com/COSMO/COSMO.owl#
## Ontology Document URL: http://micra.com/COSMO/COSMO.owl
## Using SPARQL 1.1 LOAD to get data into Virtuoso
LOAD http://micra.com/COSMO/COSMO.owl ;
WITH GRAPH http://micra.com/COSMO/COSMO.owl
INSERT {
?s http://www.w3.org/2000/01/rdf-schema#isDefinedBy
Re: [Dbpedia-discussion] A quick analysis of the classes in the DBpedia ontology
Hello, Regarding class usage, the Dutch chapter does a great job already and Magnus' query returns no results;) [1] This means that all classes are needed at the moment. Regarding the changes, I am also in favor to move forward and fix all inconsistencies. Let's start already with the obvious ones and discuss any major changes in the DBpedia meeting. Changing PopulatedPlace for instance will break many applications but if this is the way to go I am also in Cheers, Dimitris [1] http://nl.dbpedia.org/sparql?default-graph-uri=http%3A%2F%2Fnl.dbpedia .orgquery=SELECT+DISTINCT+%3Ftype+WHERE+%7B%3Ftype+a+owl%3AClass.+FILTER+NOT+EXISTS+%7B%3Fsubject+a%2Frdfs%3AsubClassOf*+%3Ftype.%7D+%7D%0D%0Aformat=text%2Fhtmltimeout=0debug=on On Sat, Apr 12, 2014 at 3:58 AM, Kingsley Idehen kide...@openlinksw.comwrote: On 4/11/14 2:12 PM, Patel-Schneider, Peter wrote: This proposal illustrates one of the major problems with the DBpedia ontology - triples check in but they never check out. Sorta, because of the misconception that SPARQL is steal Read-Only. I spend a good chunk of my day writing SPARQL 1.1 using INSERT, DELETE, and UPDATE (via INSERT and DELETE combos) to massage data, across many data spaces. Many of the problems with the DBpedia ontology require changes. Change is good. And it will work. Remember, DBpedia deploys Linked Data using a Quad Store, so the re-write rules and SPARQL queries used to perform the name-address indirection are extremely flexible. For instance, you can actually set the named graph URI scope for these SPARQL queries explicitly or via our NOT FROM NAMED GRAPH extension re., negation etc.. You can have a list of Named Graphs or excluded Named Graphs when generating the description of a DBpedia Entity URI's referent. Other problems have to do with the expressivity of the ontology. Again, its just triples to which SPARQL 1.1 patterns can be applied. In short, this is the way to truly appreciate the power of SPARQL and Linked Data. I don't think that changes here can be effected just by adding and removing bits of the ontology. You can variants of the Ontology, an alternative Ontology, it doesn't matter, the Linked Data deployment will be unaffected. Other problems have to do with the philosophy of the ontology. The philosophy of the Ontology cannot change, that's a world view of the ontology creators. That doesn't stop another ontology existing as an alternative set of context lenses into the same data. I encourage you to make your changes, or make a new ontology, whichever path you take, the end product will be useful and a showcase for perspectives sometimes overlooked due to blurred and blurry perspectives :-) We can do this, its the next stage in the natural evolution of DBpedia and the broader Linked Open Data Cloud. Note: there is zero speculation in my response. I've already done (and continue to do) a lot of this (hands on fashion) over the years, following LOD cloud initial bootstrap. BTW -- This mail was easy to write as I've just completed (literally minutes ago) a marathon session on SPARQL 1.1 which amounted to an ACL (data access policy and access control list) debugging session. Kingsley. peter On Apr 11, 2014, at 5:27 AM, Kingsley Idehen kide...@openlinksw.com wrote: On 4/11/14 5:47 AM, Magnus Knuth wrote: Hello Peter, thank you very much for your inputs. The state of the DBpedia ontology is certainly an issue. You can register at [1], ask for editing rights, and go on and make your changes. I'd also feel not quite well performing major changes or removing classes without some discussion, since it is the effort of others and it is not always clear, if somebody actually uses it. Maybe, we could organize an ontology enhancement and guidelines workshop at the next DBpedia Community Meeting in Leipzig [2]. [1]http://mappings.dbpedia.org/index.php?title=Special:UserLogin [2]http://wiki.dbpedia.org/meetings/Leipzig2014 How about the following procedure: [1] Triples relating to fixes and new relations are first created in an RDF document that 's WWW accessible [2] The documents are announced here as an invite to examine request etc.. [3] If acceptable, the Triples are added to the project [4] If unacceptable, for any reason, then in the very worst case (re. deadlocks) the Triples end up where they are on in a specific named graph in the Virtuoso instance . The real beauty of AWWW, as exemplified by Linked Data, is the ability to agree to disagree without creating inertia. Virtuoso can handle many Linked Data scenarios, and agreeing to disagree lies at the core of its design (like AWWW). Peter: we would all gladly welcome your input and contributions. The steps above will make this smooth and ultimately enlightening :-) -- Regards, Kingsley Idehen Founder CEO OpenLink Software Company Web: http://www.openlinksw.com Personal Weblog:
Re: [Dbpedia-discussion] A quick analysis of the classes in the DBpedia ontology
On Apr 11, 2014, at 11:58 AM, Kingsley Idehen kide...@openlinksw.com wrote: On 4/11/14 2:12 PM, Patel-Schneider, Peter wrote: This proposal illustrates one of the major problems with the DBpedia ontology - triples check in but they never check out. Sorta, because of the misconception that SPARQL is steal Read-Only. I spend a good chunk of my day writing SPARQL 1.1 using INSERT, DELETE, and UPDATE (via INSERT and DELETE combos) to massage data, across many data spaces. How can I use SPARQL 1.1 to change the DBpedia ontology? Many of the problems with the DBpedia ontology require changes. Change is good. And it will work. Remember, DBpedia deploys Linked Data using a Quad Store, so the re-write rules and SPARQL queries used to perform the name-address indirection are extremely flexible. For instance, you can actually set the named graph URI scope for these SPARQL queries explicitly or via our NOT FROM NAMED GRAPH extension re., negation etc.. You can have a list of Named Graphs or excluded Named Graphs when generating the description of a DBpedia Entity URI's referent. I'm not sure how any of this can be used to effect changes in the DBpedia ontology. Other problems have to do with the expressivity of the ontology. Again, its just triples to which SPARQL 1.1 patterns can be applied. In short, this is the way to truly appreciate the power of SPARQL and Linked Data. Again, I'm not sure what play SPARQL 1.1 has with respect to the expressive power of the ontology. I don't think that changes here can be effected just by adding and removing bits of the ontology. You can variants of the Ontology, an alternative Ontology, it doesn't matter, the Linked Data deployment will be unaffected. Well, sure, none of this will affect most Linked Data uses, but that's not what I'm interested in. I'm interested in using the DBpedia ontology to organize information. I could, of course, simply use a different ontology, but my hope here is that use of the DBpedia ontology in products will result in a better ontology, and that that can be shared. Other problems have to do with the philosophy of the ontology. The philosophy of the Ontology cannot change, that's a world view of the ontology creators. That doesn't stop another ontology existing as an alternative set of context lenses into the same data. It appears that the current ontology does not match the stated philosophy of the ontology. One or the other should change, and probably both. I encourage you to make your changes, or make a new ontology, whichever path you take, the end product will be useful and a showcase for perspectives sometimes overlooked due to blurred and blurry perspectives :-) I would love to make changes. There actually is a modified version of the ontology that is in use. peter We can do this, its the next stage in the natural evolution of DBpedia and the broader Linked Open Data Cloud. Note: there is zero speculation in my response. I've already done (and continue to do) a lot of this (hands on fashion) over the years, following LOD cloud initial bootstrap. BTW -- This mail was easy to write as I've just completed (literally minutes ago) a marathon session on SPARQL 1.1 which amounted to an ACL (data access policy and access control list) debugging session. Kingsley. -- Learn Graph Databases - Download FREE O'Reilly Book Graph Databases is the definitive new guide to graph databases and their applications. Written by three acclaimed leaders in the field, this first edition is now available. Download your free book today! http://p.sf.net/sfu/NeoTech ___ Dbpedia-discussion mailing list Dbpedia-discussion@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/dbpedia-discussion
Re: [Dbpedia-discussion] A quick analysis of the classes in the DBpedia ontology
Hello Peter,
thank you very much for your inputs. The state of the DBpedia ontology is
certainly an issue.
You can register at [1], ask for editing rights, and go on and make your
changes. I'd also feel not quite well performing major changes or removing
classes without some discussion, since it is the effort of others and it is not
always clear, if somebody actually uses it.
Maybe, we could organize an ontology enhancement and guidelines workshop at the
next DBpedia Community Meeting in Leipzig [2].
[1] http://mappings.dbpedia.org/index.php?title=Special:UserLogin
[2] http://wiki.dbpedia.org/meetings/Leipzig2014
On Apr 10, 2014, at 10:03:05 PM , Patel-Schneider, Peter
peter.patel-schnei...@nuance.com wrote:
On Apr 10, 2014, at 6:14 AM, Marco Fossati hell.j@gmail.com wrote:
Hi Peter,
Thank you for your detailed report.
The DBpedia ontology is (a) crowdsourced and (b) follows a data-driven
approach. Classes and properties are mainly derived from the actual data
coming from different Wikipedia chapters.
I don't think that this last is true. For example, there are about 250
classes in the ontology that do not have any instances, at least in the data
that I have examined, and more that have no instances that are not instances
of any of their subclasses. There are also a number of places where the
ontology organization does not match the information in Wikipedia.
SELECT DISTINCT ?type WHERE {?type a owl:Class. FILTER NOT EXISTS {?subject
a/rdfs:subClassOf* ?type.} }
As for the English DBpedia dataset there are 142 unused classes:
http://dbpedia.org/sparql?default-graph-uri=http%3A%2F%2Fdbpedia.orgquery=SELECT+DISTINCT+%3Ftype+WHERE+%7B%3Ftype+a+owl%3AClass.+FILTER+NOT+EXISTS+%7B%3Fsubject+a%2Frdfs%3AsubClassOf*+%3Ftype.%7D+%7Dformat=text%2Fhtmltimeout=3debug=on
As Marco said, you'd also need to consider other Language Chapters that use the
same ontology. But obviously there are some classes needless, redundant, badly
described, or just wrong.
Those are the main reasons of the issues you mentioned.
The DBpedia ontology looks much more like a unreviewed crowdsourced artifact
than an artifact that matches either the information in Wikipedia or in
DBpedia.
It would be great if you could contribute a deep analysis and detect the
inconsistencies.
Well, there are no formal inconsistencies in the DBpedia ontology, as it is
too inexpressive to have inconsistencies. All that can be done is pointing
out where the ontology does not appear to match either the normal definitions
of the categories or the definitions found by examining Wikipedia information
and differences between different parts of the ontology. I do have a list of
all the empty classes, but this information is available elsewhere.
The analysis that I sent out yesterday lists quite a number of deficiencies
in the ontology. This analysis should be good enough to serve as a start on
fixing the ontology.
In this way, we could clean the ontology up and provide rock solid
semantics.
I agree that this would be a good idea. However, this should be an iterative
process, and should not depend on a complete analysis.
As you already mention lots of examples, a brand new ontology and exact
deltas with the current one would be highly beneficial.
Well, producing a new ontology is work, and it would be nice that this work
has an effect. This is why I was wondering who else was interested in
improving the ontology.
Cheers!
peter
Best regards
Magnus
--
Magnus Knuth
Hasso-Plattner-Institut für Softwaresystemtechnik GmbH
Prof.-Dr.-Helmert-Str. 2-3
14482 Potsdam
Amtsgericht Potsdam, HRB 12184
Geschäftsführung: Prof. Dr. Christoph Meinel
tel: +49 331 5509 547
email: magnus.kn...@hpi.uni-potsdam.de
web: http://www.hpi.uni-potsdam.de/
webID: http://magnus.13mm.de/
--
Put Bad Developers to Shame
Dominate Development with Jenkins Continuous Integration
Continuously Automate Build, Test Deployment
Start a new project now. Try Jenkins in the cloud.
http://p.sf.net/sfu/13600_Cloudbees
___
Dbpedia-discussion mailing list
Dbpedia-discussion@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/dbpedia-discussion
Re: [Dbpedia-discussion] A quick analysis of the classes in the DBpedia ontology
On 4/11/14, 11:47 AM, Magnus Knuth wrote: Maybe, we could organize an ontology enhancement and guidelines workshop at the next DBpedia Community Meeting in Leipzig [2]. +1 -- Marco Fossati http://about.me/marco.fossati Twitter: @hjfocs Skype: hell_j -- Put Bad Developers to Shame Dominate Development with Jenkins Continuous Integration Continuously Automate Build, Test Deployment Start a new project now. Try Jenkins in the cloud. http://p.sf.net/sfu/13600_Cloudbees ___ Dbpedia-discussion mailing list Dbpedia-discussion@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/dbpedia-discussion
Re: [Dbpedia-discussion] A quick analysis of the classes in the DBpedia ontology
On 4/11/14 5:47 AM, Magnus Knuth wrote: Hello Peter, thank you very much for your inputs. The state of the DBpedia ontology is certainly an issue. You can register at [1], ask for editing rights, and go on and make your changes. I'd also feel not quite well performing major changes or removing classes without some discussion, since it is the effort of others and it is not always clear, if somebody actually uses it. Maybe, we could organize an ontology enhancement and guidelines workshop at the next DBpedia Community Meeting in Leipzig [2]. [1]http://mappings.dbpedia.org/index.php?title=Special:UserLogin [2]http://wiki.dbpedia.org/meetings/Leipzig2014 How about the following procedure: [1] Triples relating to fixes and new relations are first created in an RDF document that 's WWW accessible [2] The documents are announced here as an invite to examine request etc.. [3] If acceptable, the Triples are added to the project [4] If unacceptable, for any reason, then in the very worst case (re. deadlocks) the Triples end up where they are on in a specific named graph in the Virtuoso instance . The real beauty of AWWW, as exemplified by Linked Data, is the ability to agree to disagree without creating inertia. Virtuoso can handle many Linked Data scenarios, and agreeing to disagree lies at the core of its design (like AWWW). Peter: we would all gladly welcome your input and contributions. The steps above will make this smooth and ultimately enlightening :-) -- Regards, Kingsley Idehen Founder CEO OpenLink Software Company Web: http://www.openlinksw.com Personal Weblog: http://www.openlinksw.com/blog/~kidehen Twitter Profile: https://twitter.com/kidehen Google+ Profile: https://plus.google.com/+KingsleyIdehen/about LinkedIn Profile: http://www.linkedin.com/in/kidehen smime.p7s Description: S/MIME Cryptographic Signature -- Put Bad Developers to Shame Dominate Development with Jenkins Continuous Integration Continuously Automate Build, Test Deployment Start a new project now. Try Jenkins in the cloud. http://p.sf.net/sfu/13600_Cloudbees___ Dbpedia-discussion mailing list Dbpedia-discussion@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/dbpedia-discussion
Re: [Dbpedia-discussion] A quick analysis of the classes in the DBpedia ontology
On 4/11/14 2:12 PM, Patel-Schneider, Peter wrote: This proposal illustrates one of the major problems with the DBpedia ontology - triples check in but they never check out. Sorta, because of the misconception that SPARQL is steal Read-Only. I spend a good chunk of my day writing SPARQL 1.1 using INSERT, DELETE, and UPDATE (via INSERT and DELETE combos) to massage data, across many data spaces. Many of the problems with the DBpedia ontology require changes. Change is good. And it will work. Remember, DBpedia deploys Linked Data using a Quad Store, so the re-write rules and SPARQL queries used to perform the name-address indirection are extremely flexible. For instance, you can actually set the named graph URI scope for these SPARQL queries explicitly or via our NOT FROM NAMED GRAPH extension re., negation etc.. You can have a list of Named Graphs or excluded Named Graphs when generating the description of a DBpedia Entity URI's referent. Other problems have to do with the expressivity of the ontology. Again, its just triples to which SPARQL 1.1 patterns can be applied. In short, this is the way to truly appreciate the power of SPARQL and Linked Data. I don't think that changes here can be effected just by adding and removing bits of the ontology. You can variants of the Ontology, an alternative Ontology, it doesn't matter, the Linked Data deployment will be unaffected. Other problems have to do with the philosophy of the ontology. The philosophy of the Ontology cannot change, that's a world view of the ontology creators. That doesn't stop another ontology existing as an alternative set of context lenses into the same data. I encourage you to make your changes, or make a new ontology, whichever path you take, the end product will be useful and a showcase for perspectives sometimes overlooked due to blurred and blurry perspectives :-) We can do this, its the next stage in the natural evolution of DBpedia and the broader Linked Open Data Cloud. Note: there is zero speculation in my response. I've already done (and continue to do) a lot of this (hands on fashion) over the years, following LOD cloud initial bootstrap. BTW -- This mail was easy to write as I've just completed (literally minutes ago) a marathon session on SPARQL 1.1 which amounted to an ACL (data access policy and access control list) debugging session. Kingsley. peter On Apr 11, 2014, at 5:27 AM, Kingsley Idehen kide...@openlinksw.com wrote: On 4/11/14 5:47 AM, Magnus Knuth wrote: Hello Peter, thank you very much for your inputs. The state of the DBpedia ontology is certainly an issue. You can register at [1], ask for editing rights, and go on and make your changes. I'd also feel not quite well performing major changes or removing classes without some discussion, since it is the effort of others and it is not always clear, if somebody actually uses it. Maybe, we could organize an ontology enhancement and guidelines workshop at the next DBpedia Community Meeting in Leipzig [2]. [1]http://mappings.dbpedia.org/index.php?title=Special:UserLogin [2]http://wiki.dbpedia.org/meetings/Leipzig2014 How about the following procedure: [1] Triples relating to fixes and new relations are first created in an RDF document that 's WWW accessible [2] The documents are announced here as an invite to examine request etc.. [3] If acceptable, the Triples are added to the project [4] If unacceptable, for any reason, then in the very worst case (re. deadlocks) the Triples end up where they are on in a specific named graph in the Virtuoso instance . The real beauty of AWWW, as exemplified by Linked Data, is the ability to agree to disagree without creating inertia. Virtuoso can handle many Linked Data scenarios, and agreeing to disagree lies at the core of its design (like AWWW). Peter: we would all gladly welcome your input and contributions. The steps above will make this smooth and ultimately enlightening :-) -- Regards, Kingsley Idehen Founder CEO OpenLink Software Company Web: http://www.openlinksw.com Personal Weblog: http://www.openlinksw.com/blog/~kidehen Twitter Profile: https://twitter.com/kidehen Google+ Profile: https://plus.google.com/+KingsleyIdehen/about LinkedIn Profile: http://www.linkedin.com/in/kidehen -- Put Bad Developers to Shame Dominate Development with Jenkins Continuous Integration Continuously Automate Build, Test Deployment Start a new project now. Try Jenkins in the cloud. http://p.sf.net/sfu/13600_Cloudbees___ Dbpedia-discussion mailing list Dbpedia-discussion@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/dbpedia-discussion -- Regards, Kingsley Idehen Founder CEO OpenLink Software Company Web: http://www.openlinksw.com Personal Weblog: http://www.openlinksw.com/blog/~kidehen Twitter Profile:
Re: [Dbpedia-discussion] A quick analysis of the classes in the DBpedia ontology
On Apr 10, 2014, at 6:14 AM, Marco Fossati hell.j@gmail.com wrote: Hi Peter, Thank you for your detailed report. The DBpedia ontology is (a) crowdsourced and (b) follows a data-driven approach. Classes and properties are mainly derived from the actual data coming from different Wikipedia chapters. I don't think that this last is true. For example, there are about 250 classes in the ontology that do not have any instances, at least in the data that I have examined, and more that have no instances that are not instances of any of their subclasses. There are also a number of places where the ontology organization does not match the information in Wikipedia. Those are the main reasons of the issues you mentioned. The DBpedia ontology looks much more like a unreviewed crowdsourced artifact than an artifact that matches either the information in Wikipedia or in DBpedia. It would be great if you could contribute a deep analysis and detect the inconsistencies. Well, there are no formal inconsistencies in the DBpedia ontology, as it is too inexpressive to have inconsistencies. All that can be done is pointing out where the ontology does not appear to match either the normal definitions of the categories or the definitions found by examining Wikipedia information and differences between different parts of the ontology. I do have a list of all the empty classes, but this information is available elsewhere. The analysis that I sent out yesterday lists quite a number of deficiencies in the ontology. This analysis should be good enough to serve as a start on fixing the ontology. In this way, we could clean the ontology up and provide rock solid semantics. I agree that this would be a good idea. However, this should be an iterative process, and should not depend on a complete analysis. As you already mention lots of examples, a brand new ontology and exact deltas with the current one would be highly beneficial. Well, producing a new ontology is work, and it would be nice that this work has an effect. This is why I was wondering who else was interested in improving the ontology. Cheers! peter -- Put Bad Developers to Shame Dominate Development with Jenkins Continuous Integration Continuously Automate Build, Test Deployment Start a new project now. Try Jenkins in the cloud. http://p.sf.net/sfu/13600_Cloudbees ___ Dbpedia-discussion mailing list Dbpedia-discussion@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/dbpedia-discussion
Re: [Dbpedia-discussion] A quick analysis of the classes in the DBpedia ontology
Hi, I think that Peter has good reasons to complain on the current status of the DBpedia ontology :) But besides debatable choices on names, I’d concentrate on the main issue, which is the data-grounding of the ontology. The major example is that there is no systematic checking of the relation between domain/ranges and the way properties are actually used (although the situation has improved in 3.9). An organization of classes should be based firstly on how properties are used, after all we are talking primarily linked data, not taxonomies. Some research has been done on this, and I will contribute some literature to the community if any activity starts on reconciling crowd-sourcing, automatic extraction of data, and good practices of ontology design. Best Aldo On Apr 10, 2014, at 10:03:05 PM , Patel-Schneider, Peter peter.patel-schnei...@nuance.com wrote: On Apr 10, 2014, at 6:14 AM, Marco Fossati hell.j@gmail.com wrote: Hi Peter, Thank you for your detailed report. The DBpedia ontology is (a) crowdsourced and (b) follows a data-driven approach. Classes and properties are mainly derived from the actual data coming from different Wikipedia chapters. I don't think that this last is true. For example, there are about 250 classes in the ontology that do not have any instances, at least in the data that I have examined, and more that have no instances that are not instances of any of their subclasses. There are also a number of places where the ontology organization does not match the information in Wikipedia. Those are the main reasons of the issues you mentioned. The DBpedia ontology looks much more like a unreviewed crowdsourced artifact than an artifact that matches either the information in Wikipedia or in DBpedia. It would be great if you could contribute a deep analysis and detect the inconsistencies. Well, there are no formal inconsistencies in the DBpedia ontology, as it is too inexpressive to have inconsistencies. All that can be done is pointing out where the ontology does not appear to match either the normal definitions of the categories or the definitions found by examining Wikipedia information and differences between different parts of the ontology. I do have a list of all the empty classes, but this information is available elsewhere. The analysis that I sent out yesterday lists quite a number of deficiencies in the ontology. This analysis should be good enough to serve as a start on fixing the ontology. In this way, we could clean the ontology up and provide rock solid semantics. I agree that this would be a good idea. However, this should be an iterative process, and should not depend on a complete analysis. As you already mention lots of examples, a brand new ontology and exact deltas with the current one would be highly beneficial. Well, producing a new ontology is work, and it would be nice that this work has an effect. This is why I was wondering who else was interested in improving the ontology. Cheers! peter -- Put Bad Developers to Shame Dominate Development with Jenkins Continuous Integration Continuously Automate Build, Test Deployment Start a new project now. Try Jenkins in the cloud. http://p.sf.net/sfu/13600_Cloudbees ___ Dbpedia-discussion mailing list Dbpedia-discussion@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/dbpedia-discussion -- Put Bad Developers to Shame Dominate Development with Jenkins Continuous Integration Continuously Automate Build, Test Deployment Start a new project now. Try Jenkins in the cloud. http://p.sf.net/sfu/13600_Cloudbees ___ Dbpedia-discussion mailing list Dbpedia-discussion@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/dbpedia-discussion
[Dbpedia-discussion] A quick analysis of the classes in the DBpedia ontology
I did a quick analysis of the classes in the DBpedia ontology and found quite a few issues that I think need attention. - Many classes have no instances. Each of these empty classes should be examined to see whether they should be removed or modified. - The sports-related groupings are differentially populated, differentially organized, and unaxiomatized. These groupings should be regularized and minimal axiomatizations provided for them. For example, there would be classes for Basketball under at least SportsLeague, SportsTeam, Coach, and SportsEvent each defined as the restriction of the grouping elements related to Basketball. The sports groupings include Sport (which is special), SportsLeague, SportsTeam, Athlete, Coach, SportsTeamMember, SportsManager, SportsEvent, SportFacility, SportCompetitionResult, SportsSeason, and Tournament. - Numerous stated inclusion relationships are not correct when considering the normal definition of the class names. Each of these should be examined and either descriptions of the classes that support the inclusion relationship be provided or the relationship itself modified. For example, instances of the RecordOffice class do not appear to be non-profit organizations. Some other examples of questionable or outright incorrect subclasses here are TermOfOffice, BackScene, ChessPlayer, PokerPlayer, TeamMember, Saint, FictionalCharacter, MythologicalFigure, OrganisationMember, Religious, Baronet, Medician, Professor, Embryology, Lymph, Constellation, Galaxy, ElectionDiagram, Olympics, OlympicEvent, ControlledDesignationOfOriginWine, PublicServiceInput, PublicServiceOutput, and ProgrammingLanguage. - Some class relationships are missing. For example, TeamMember is unrelated to SportsTeamMember even though they are both supposed to be members of athletic teams. Some other examples of missing relationships are between BullFighter and Bullfighter, between Host and TelevisionHost, and between Comic and Comics. The missing relationships should be provided or the classes merged. - Place is a rather unnatural union. It should either be removed or better organized. - There are quite a few subclasses of Building that are not truely buildings, including AmusementParkAttraction, Casino, Factory, Hotel, MilitaryStructure, Abbey and the other religious places of worship, Restaurant, ShoppingMall, and Venue. Similarly, there are a number of subclasses of ArchitecturalStructure that may not be architectural structures, including Garden, PublicTransitSystem, and Park. There are a few subclasses of NaturalPlace that are not necessarily natural places, including Canal, and even Lake. These classes should be moved up in the ontology. - The subclasses of Species are not collections of species. The subclasses should either be modified or moved elsewhere in the ontology. - The normal definition of PopulatedPlace is much too narrow to encompass all its subclasses. A new general class should be created to encompass the subclasses and PopulatedPlace be modified as necessary. - There are a number of strange top-level or second-level classes. These classes should be examined to ensure that they make sense. Many of these classes appear to be somehow related to measurements, including Altitude, Area, Blazon, ChartsPlacement, Demographics, Depth, GrossDomesticProduct, GrossDomesticProductPerCapita, HumanDevelopmentIndex, Population, Sales, Statistics, and Tax. Other strange classes include LifeCycleEvent, Imdb, Listen, PenaltyShootOut, PersonFunction, PoliticalFunction, Profession, TopicalConcept, Type, and YearInSpaceflight. Even if I had editing rights to the ontology I think that the fixes I have outlined above go beyond what should be done without some discussion. Comments? peter -- Put Bad Developers to Shame Dominate Development with Jenkins Continuous Integration Continuously Automate Build, Test Deployment Start a new project now. Try Jenkins in the cloud. http://p.sf.net/sfu/13600_Cloudbees ___ Dbpedia-discussion mailing list Dbpedia-discussion@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/dbpedia-discussion
