anuzzolese / oke-challenge

19 stars 8 forks source link

Issue on sameAs link in task-1 training set #10

Closed jplu closed 9 years ago

jplu commented 9 years ago

Hi,

I saw a new bug in the dataset:

<http://www.ontologydesignpatterns.org/data/oke-challenge/task-1/St._Mark's_School>
        a           dul:Organization , owl:Individual ;
        rdfs:label  "St. Mark's School"@en ;
        owl:sameAs  dbpedia: .

Should be:

<http://www.ontologydesignpatterns.org/data/oke-challenge/task-1/St._Mark's_School>
        a           dul:Organization , owl:Individual ;
        rdfs:label  "St. Mark's School"@en .

Unless there is a DBpedia resource, but I didnt find it.

Cheers.

giusepperizzo commented 9 years ago

curl http://dbpedia.org/resource/St._Mark\'s_School -> 303 chromium http://dbpedia.org/resource/St._Mark\'s_School -> "No further information is available. (The requested entity is unknown)"

But if you describe it via SPARQL: describe http://dbpedia.org/resource/St._Mark's_School

it exists. So far, it's not the 1st case I've been experiencing.

@anuzzolese @AnLiGentile will you still consider this as a valid disambiguation resource?

MichaelRoeder commented 9 years ago

Hi all,

just my two cents: there seem to be more than one single St. Mark's Schools. The sentence starts with

At St. Mark's School in Southborough, Massachusetts, ...

This school should be reachable using the URI http://dbpedia.org/resource/St._Mark's_School_(Massachusetts), since it is referenced by the city http://dbpedia.org/resource/Southborough,_Massachusetts using this URI. However, this resource does not seem to exist.

Cheers Michael

rtroncy commented 9 years ago

Actuallly, http://dbpedia.org/resource/St._Mark's_School_(Massachusetts) does exist, but don't trust your browser! There is a problem in how Openlink dereferences URIs that contain characters such as '. Do instead a:

DESCRIBE <http://dbpedia.org/resource/St._Mark's_School_(Massachusetts)>

and you will get the description of this resource

MichaelRoeder commented 9 years ago

Yes, the SPARQL query is working. Thanks for this hint! Thus, from my point of view, it should be

<http://www.ontologydesignpatterns.org/data/oke-challenge/task-1/St._Mark's_School>
        a           dul:Organization , owl:Individual ;
        rdfs:label  "St. Mark's School"@en ;
        owl:sameAs  dbpedia:St._Mark's_School_(Massachusetts) .
anuzzolese commented 9 years ago

Hi all,

thanks for your precious hints. dbpedia:St._Mark’s_School is actually a redirect, hence dbpedia:St._Mark'sSchool(Massachusetts) is correct. This is going to be fixed in the data.

Andrea

On 31 Mar 2015, at 11:48, Michael Röder notifications@github.com wrote:

Yes, the SPARQL query is working. Thanks for this hint! Thus, from my point of view, it should be

http://www.ontologydesignpatterns.org/data/oke-challenge/task-1/St._Mark's_School a dul:Organization , owl:Individual ; rdfs:label "St. Mark's School"@en ; owl:sameAs dbpedia:St._Mark'sSchool(Massachusetts) . — Reply to this email directly or view it on GitHub https://github.com/anuzzolese/oke-challenge/issues/10#issuecomment-88018084.

rtroncy commented 9 years ago

@anuzzolese Are you saying we should expect a new training dataset anytime soon? Will you provide as well a diff log of what has been fixed?

anuzzolese commented 9 years ago

@rtroncy the training set has been already updated. In such an update

http://www.ontologydesignpatterns.org/data/oke-challenge/task-1/St._Mark's_School a dul:Organization , owl:Individual ; rdfs:label "St. Mark's School"@en ; owl:sameAs dbpedia: .

has been changed to

http://www.ontologydesignpatterns.org/data/oke-challenge/task-1/St._Mark's_School a dul:Organization , owl:Individual ; rdfs:label "St. Mark's School"@en ; owl:sameAs dbpedia:St._Mark'sSchool(Massachusetts) .