Instrument taxonomy
Yves Raimond
I was just checking out the instrument taxonomy that Ivan derived from
Musicbraiz
http://purl.org/ontology/mo/mit/
For each instrument-related concept in it, there is the following statement:
<skos:inScheme rdf:resource="http://purl.org/ontology/mo/instruments#Musical_instruments"/>
which I don't really understand? wouldn't it be something like
<skos:inScheme rdf:resource=""/> ?
Moreover, I was thinking that it'd perhaps be better if the top
concept was mo:Instrument, and then every instrument is derived from
that.
Like:
<> skos:hasTopConcept mo:Instrument.
mo:Instrument
skos:narrower mit:Wind_Instruments;
skos:narrower mit:String_instruments;
skos:narrower mit:Percussion_instruments.
etc...
Cheers!
y
Tom Keays
Musical Instrument Taxonomy posting and your post was one of the first
I saw in the archive.
http://ivanherman.wordpress.com/2007/03/30/taxonomy-of-musical-instruments-in-skos/
You are raising some of the same questions I had about the
relationship MIT has to Music Ontology.
On Mar 5, 10:32 am, "Yves Raimond" <yves.raim...@gmail.com> wrote:
> I was just checking out the instrument taxonomy that Ivan derived from
> Musicbraizhttp://purl.org/ontology/mo/mit/
In MIT, http://purl.org/ontology/mo/mit/Bagpipe would 303 redirect to
http://www.ivan-herman.net/Misc/2007/Mu.rdf#Bagpipe
However, the URI,
http://purl.org/ontology/mo/instruments#Musical_instruments doesn't do
what you would expect. Instead, it redirects to
http://musicontology.com/musicontology.rdfs#instruments.
> For each instrument-related concept in it, there is the following statement:
> <skos:inScheme rdf:resource="http://purl.org/ontology/mo/instruments#Musical_instruments"/>
Note, Music Ontology has an "Instrument" class rather than a
"Musical_instruments" class. I.E.,
<rdfs:Class rdf:about="http://purl.org/ontology/mo/Instrument"
rdfs:label="Instrument" vs:term_status="stable" mo:level="2">
<rdfs:comment>
Any of various devices or contrivances that can be used to
produce musical tones or sound.
Any taxonomy can be used to subsume this concept. The default
one is one extracted by Ivan Herman from the Musicbrainz instrument
taxonomy, conforming to SKOS. This concept holds a seeAlso link
towards this taxonomy.
</rdfs:comment>
<rdfs:isDefinedBy rdf:resource="http://purl.org/ontology/mo/"/>
<rdf:type rdf:resource="http://www.w3.org/2002/07/owl#Class"/>
<rdfs:seeAlso rdf:resource="http://purl.org/ontology/mo/mit#"/>
</rdfs:Class>
Therefore, in the Music Ontology schema, the proper label would be
"Instrument".
That is, the URI above really should be
http://purl.org/ontology/mo/Instrument and it would redirect to
http://musicontology.com/musicontology.rdfs#Instrument.
So, to my mind, that needs to get fixed in MIT.
> which I don't really understand? wouldn't it be something like
> <skos:inScheme rdf:resource=""/> ?
Why? I don't really know SKOS all that well yet, but I don't see the
point of an empty URI here. It ought to refer back to MO. Just not the
way that has been done.
> Moreover, I was thinking that it'd perhaps be better if the top
> concept was mo:Instrument, and then every instrument is derived from
> that.
> Like:
>
> <> skos:hasTopConcept mo:Instrument.
>
> mo:Instrument
> skos:narrower mit:Wind_Instruments;
> skos:narrower mit:String_instruments;
> skos:narrower mit:Percussion_instruments.
> etc...
The problem that I see here is that MO is not defined as a SKOS, so I
don't believe you can declare mo:Instrument to be a top concept.
MIT for some reason fails to name its ConceptScheme and that may be
what's bothering you.
Here's how it is being done now:
<skos:ConceptScheme rdf:about="">
... stuff omitted ...
<skos:hasTopConcept
rdf:resource="http://purl.org/ontology/mo/mit#Wind_instruments"/>
<skos:hasTopConcept
rdf:resource="http://purl.org/ontology/mo/mit#String_instruments"/>
<skos:hasTopConcept
rdf:resource="http://purl.org/ontology/mo/mit#Percussion_instruments"/>
<skos:hasTopConcept
rdf:resource="http://purl.org/ontology/mo/mit#Electronic_instruments"/>
<skos:hasTopConcept
rdf:resource="http://purl.org/ontology/mo/mit#Other_instruments"/>
</skos:ConceptScheme>
Perhaps that's why it goes on to define the "inScheme" predicate in
terms of MO "Instrument"
<skos:Concept rdf:about="http://purl.org/ontology/mo/mit#Wind_instruments">
<skos:prefLabel>Wind instruments</skos:prefLabel>
<skos:inScheme
rdf:resource="http://purl.org/ontology/mo/instruments#Musical_instruments"/>
<skos:narrower>
<skos:Concept rdf:about="http://purl.org/ontology/mo/mit#Woodwind">
<skos:prefLabel>Woodwind</skos:prefLabel>
<skos:inScheme
rdf:resource="http://purl.org/ontology/mo/instruments#Musical_instruments"/>
<skos:narrower>
... stuff omitted ...
</skos:narrower>
</skos:Concept>
</skos:narrower>
</skos:Concept>
First, correct all occurences of:
<skos:inScheme rdf:resource="http://purl.org/ontology/mo/instruments#Musical_instruments"/>
to:
<skos:inScheme rdf:resource="http://purl.org/ontology/mo/Instrument"/>
Then, I guess the question that remains is why it wasn't done more
like the examples from the SKOS spec --
http://www.w3.org/TR/swbp-skos-core-spec/#concept-schemes
My suggestion would be more like this:
<mit:Instruments> rdf:type skos:ConceptScheme ;
rdfs:seeAlso <mo:Instrument> ;
// can't use owl:sameAs, owl:equivalentClass or owl:equivalentProperty
skos:hasTopConcept <mit:Wind_instruments> ;
skos:hasTopConcept <mit:String_instruments> ;
skos:hasTopConcept <mit:Percussion_instruments> ;
skos:hasTopConcept <mit:Electronic_instruments> ;
skos:hasTopConcept <mit:Other_instruments> .
<mit:Wind_instruments> skos:inScheme <mit:Instruments> .
<mit:String_instruments> skos:inScheme <mit:Instruments> .
<mit:Percussion_instruments> skos:inScheme <mit:Instruments> .
<mit:Electronic_instruments> skos:inScheme <mit:Instruments> .
<mit:Other_instruments> skos:inScheme <mit:Instruments> .
Tom
Yves Raimond
> Nice coincidence. I discovered this group today via Ivan Herman's
Welcome on board :-)
> Musical Instrument Taxonomy posting and your post was one of the first
> I saw in the archive.
>
> http://ivanherman.wordpress.com/2007/03/30/taxonomy-of-musical-instruments-in-skos/
>
> You are raising some of the same questions I had about the
> relationship MIT has to Music Ontology.
>
>
> On Mar 5, 10:32 am, "Yves Raimond" <yves.raim...@gmail.com> wrote:
> > I was just checking out the instrument taxonomy that Ivan derived from
> > Musicbraizhttp://purl.org/ontology/mo/mit/
>
> In MIT, http://purl.org/ontology/mo/mit/Bagpipe would 303 redirect to
> http://www.ivan-herman.net/Misc/2007/Mu.rdf#Bagpipe
>
> However, the URI,
> http://purl.org/ontology/mo/instruments#Musical_instruments doesn't do
> what you would expect. Instead, it redirects to
> http://musicontology.com/musicontology.rdfs#instruments.
Hmm... yes, this URI is broken, I actually wonder where it comes from
and if it was once dereferencable. An instrument in MO is identified
by http://purl.org/ontology/mo/Instrument
>
>
> > For each instrument-related concept in it, there is the following statement:
> > <skos:inScheme rdf:resource="http://purl.org/ontology/mo/instruments#Musical_instruments"/>
>
> Note, Music Ontology has an "Instrument" class rather than a
> "Musical_instruments" class. I.E.,
>
> <rdfs:Class rdf:about="http://purl.org/ontology/mo/Instrument"
> rdfs:label="Instrument" vs:term_status="stable" mo:level="2">
> <rdfs:comment>
> Any of various devices or contrivances that can be used to
> produce musical tones or sound.
> Any taxonomy can be used to subsume this concept. The default
> one is one extracted by Ivan Herman from the Musicbrainz instrument
> taxonomy, conforming to SKOS. This concept holds a seeAlso link
> towards this taxonomy.
> </rdfs:comment>
> <rdfs:isDefinedBy rdf:resource="http://purl.org/ontology/mo/"/>
> <rdf:type rdf:resource="http://www.w3.org/2002/07/owl#Class"/>
> <rdfs:seeAlso rdf:resource="http://purl.org/ontology/mo/mit#"/>
> </rdfs:Class>
>
> Therefore, in the Music Ontology schema, the proper label would be
> "Instrument".
>
> That is, the URI above really should be
> http://purl.org/ontology/mo/Instrument and it would redirect to
> http://musicontology.com/musicontology.rdfs#Instrument.
>
> So, to my mind, that needs to get fixed in MIT.
>
Yes, I'll give it a try and send the corresponding patch to Ivan soon.
>
> > which I don't really understand? wouldn't it be something like
> > <skos:inScheme rdf:resource=""/> ?
>
> Why? I don't really know SKOS all that well yet, but I don't see the
> point of an empty URI here. It ought to refer back to MO. Just not the
> way that has been done.
>
Hmm... I put an empty URI here to refer to the current URI (that is,
http://purl.org/ontology/mo/mit), which could identify the SKOS
concept scheme itself? Although it might indeed be broken.
The main idea was to define the concept scheme in MIT, and this
concept scheme could have its top-level concept in MO (mo:Instrument),
but indeed, it can cause some troubles (see below).
>
> > Moreover, I was thinking that it'd perhaps be better if the top
> > concept was mo:Instrument, and then every instrument is derived from
> > that.
> > Like:
> >
> > <> skos:hasTopConcept mo:Instrument.
> >
> > mo:Instrument
> > skos:narrower mit:Wind_Instruments;
> > skos:narrower mit:String_instruments;
> > skos:narrower mit:Percussion_instruments.
> > etc...
>
> The problem that I see here is that MO is not defined as a SKOS, so I
> don't believe you can declare mo:Instrument to be a top concept.
>
Yes, I do wonder whether that's right too... mo:Instrument is not a
skos:Concept.
Hmm... I agree with that, but don't get the link between the concept
scheme and the instrument concept within MO (rdfs:seeAlso doesn't seem
appropriate here).
Perhaps another solution would be to make MIT completely stand-alone
(not refering to mo:Instrument at all?), and add skos:Concept to the
range of mo:instrument... But it looks a bit dirty to me.
To sum up, I think the biggest issue is the fact that mo:Instrument is
a owl:Class in MO, and not a skos:Concept neither a ConceptScheme. I
don't think it'd be feasable to make it that way, though (I might want
to refer to my very own flute, for example :-) ).
Cheers!
y