Questions on MusicBrainz data

[Tom Morris]

These questions apparently got lost in the thread on the update, so
pulling them into their own thread...

- how are the sameAs links for the MusicBrainz data set calculated?
Is it meant to include the transitive closure of all sameAs properties
on the server?

- What are the three copies of linked MusicBrainz data? Ref:

On Fri, Aug 13, 2010 at 9:43 AM, Ian Davis <m...@iandavis.com> wrote:
> There are 3 different expressions of the MB data to my knowledge.

Tom

[Richard Cyganiak]

On 18 Aug 2010, at 21:53, Tom Morris wrote:
> - What are the three copies of linked MusicBrainz data? Ref:

Ian Davis'
http://musicbrainz.dataincubator.org/

Yves Raimond's
http://dbtune.org/musicbrainz/

Fred Giasson's
http://fgiasson.com/blog/index.php/2007/05/22/browsing-musicbrainzs-dataset-via-uri-dereferencing/

Best,
Richard

>
> On Fri, Aug 13, 2010 at 9:43 AM, Ian Davis <m...@iandavis.com> wrote:
>> There are 3 different expressions of the MB data to my knowledge.
>
> Tom
>

> --
> You received this message because you are subscribed to the Google
> Groups "Data Incubator" group.
> To post to this group, send email to datain...@googlegroups.com.
> To unsubscribe from this group, send email to dataincubato...@googlegroups.com
> .
> For more options, visit this group at http://groups.google.com/group/dataincubator?hl=en
> .
>

[Ian Davis]

On Wed, Aug 18, 2010 at 9:53 PM, Tom Morris <tfmo...@gmail.com> wrote:
> These questions apparently got lost in the thread on the update, so
> pulling them into their own thread...
>
> - how are the sameAs links for the MusicBrainz data set calculated?
> Is it meant to include the transitive closure of all sameAs properties
> on the server?

No. The triple store it is being served from does not have reasoning
or smushing enabled.

It's just the physical triples I generated from my python scripts.

Ian

[Tom Morris]

Thanks guys.

On Wed, Aug 18, 2010 at 6:23 PM, Richard Cyganiak <ric...@cyganiak.de> wrote:
> On 18 Aug 2010, at 21:53, Tom Morris wrote:
>>
>> - What are the three copies of linked MusicBrainz data?  Ref:
>
> Ian Davis'
> http://musicbrainz.dataincubator.org/
>
> Yves Raimond's
> http://dbtune.org/musicbrainz/
>
> Fred Giasson's
> http://fgiasson.com/blog/index.php/2007/05/22/browsing-musicbrainzs-dataset-via-uri-dereferencing/

Thanks for the links. I'd heard about DBtune (but forgotten it), but
not Fred's version.

So I guess the total is four or five if you consider data sets that
include MusicBrainz as part of a greater whole like the BBC or
Freebase -- with the MusicBrainz native version coming as an
additional one in the future. I wonder if anyone has compared them
all for freshness, quality, link-density, ontologies used, etc.

Tom

[Kurt J]

I'm quite sure that Ian's is the "freshest" by a fairly safe margin.
It seems to be the only one that deals with MusicBrainz Next
Generation Schema and Advanced Relationships.

Note to Ian - http://musicbrainz.dataincubator.org/.html indicates the
dct:source is http://musicbrainz.org/products/database/getdump.html?file=mbdump.tar.bz2
which i believe is inaccurate as that would be the old DB schema, not
NGS.

I know that the DBTune endpoint is based on a rather old MB DB dump
(well over a year old). It's pretty likely we'll update the DBTune
endpoint as part of the LinkedBrainz project - we will be exploring
various RDB 2 RDF solutions (aka RDF views) to run directly from
MusicBrainz. DBtune/MusicBrainz will serve as a testing ground. The
MusicBrainz dev team is pretty concerned about placing additional load
on their DB server so we have to be careful to respect that concern.

For now, we're implmenting RDFa in some of the core entity pages which
will hopefully find its way into the first major NGS release. You can
follow our progress at

http://linkedbrainz.c4dmpresents.org/

and discussions about RDf in MusicBrainz have taken place on the Music
Ontology specification list and will continue to do so.

http://groups.google.com/group/music-ontology-specification-group

thanks,
Kurt J