Re: last.fm events RDFizing

[Kingsley Idehen]

Keith Alexander wrote:

Hi

http://lastfm.rdfize.com/

RDFizes the last.fm events API and links to  dbtune.org, the bbc, and 
dbpedia.


The voiD document describing it is: http://lastfm.rdfize.com/meta.n3

Thanks,

Keith Alexander



Keith,

Hopefully, you've seen my Dataincubator posting re. all the xslt 
sytlesheets [1] that drive the 30+ sources we RDFize :-) That said, I'll 
take a look at your stuff and see how it compares to ours re. fidelity etc..


Others: I've just reiterated to the dataincubator community, by way of 
source directory exposure, that our Sponger Cartridges (Extractors or 
Mappers) have been Open Source since inception (circa. 2006). Thus, it 
would be nice for the community to leverage the base we have en route to 
generating a more coherent Web of Dynamically Linked Data via RDFization 
(or Wrapping).  Tweaking  and fixing existing mappers trumps writing 
from scratch :-)


Links:

1. http://demo.openlinksw.com/DAV/VAD/rdf_mappers/xslt/  -- all the 
extraction and mapping xslt resources


--


Regards,

Kingsley Idehen   Weblog: http://www.openlinksw.com/blog/~kidehen
President  CEO 
OpenLink Software Web: http://www.openlinksw.com

Re: last.fm events RDFizing

[Yves Raimond]

Hello!


 Hopefully we can make one coherent Linked Data Space via xslt at:
 http://demo.openlinksw.com/DAV/VAD/rdf_mappers/xslt/lastfm2rdf.xsl

 Example:
 http://linkeddata.uriburner.com/about/html/http://linkeddata.uriburner.com/about/rdf/http://www.last.fm/music/Con+Funk+Shun%23this

 As per earlier post, all the xslt components of our sponger cartridges are
 at:
 http://demo.openlinksw.com/DAV/VAD/rdf_mappers/xslt/

 Let's tweak rather that reinvent the wheel, everyone is ultimately time
 challenged :-)


Wow! That is great!! Perhaps you should contact Last.FM to ask them to
link to these stylesheets as GRDDL transformations (this is also valid
for the other services RDFised by the stylesheets you mentioned
above)?
However, I think these three RDFisers are complementary.
For example,
http://linkeddata.uriburner.com/about/html/http://linkeddata.uriburner.com/about/rdf/http://www.last.fm/event/958377
seems to RDFise the microformatted event, whereas Keith's description
is much richer and especially holds links to other services.
Same thing for profile pages, which give less information and less
links than Keith's service (dealing with events) the corresponding
dbtune service (dealing with foaf:knows and last scrobbles).

Cheers,
y

Re: last.fm events RDFizing

[Kingsley Idehen]

Yves Raimond wrote:

Hello!

  

Hopefully we can make one coherent Linked Data Space via xslt at:
http://demo.openlinksw.com/DAV/VAD/rdf_mappers/xslt/lastfm2rdf.xsl

Example:
http://linkeddata.uriburner.com/about/html/http://linkeddata.uriburner.com/about/rdf/http://www.last.fm/music/Con+Funk+Shun%23this

As per earlier post, all the xslt components of our sponger cartridges are
at:
http://demo.openlinksw.com/DAV/VAD/rdf_mappers/xslt/

Let's tweak rather that reinvent the wheel, everyone is ultimately time
challenged :-)




Wow! That is great!! Perhaps you should contact Last.FM to ask them to
link to these stylesheets as GRDDL transformations (this is also valid
for the other services RDFised by the stylesheets you mentioned
above)?
  

Yes, they all use a consistent pattern re. proxy/wrapper linked data URIs.

Re. last.fm folks, if me combine efforts and produce meshes with good 
domain fidelity, we can simply use the owl:shameAs pattern to get them 
on board :-)



However, I think these three RDFisers are complementary.
For example,
http://linkeddata.uriburner.com/about/html/http://linkeddata.uriburner.com/about/rdf/http://www.last.fm/event/958377
seems to RDFise the microformatted event, whereas Keith's description
is much richer and especially holds links to other services.
  
No, its just little bugs in the xslt. This cartridge is supposed to make 
a data space comprised of many things:


Artist Descriptions, Events, Genre's etc..

Our problem is that we've built 30+ cartridges and counting, and this 
leads to subtle drops in fidelity when you look at some of the graphs 
with domain specificity. Hence my callout for everyone to put heads 
together since we have all the pieces, but are missing the critical 
essence: working together rather than past one another en route to the 
same destination :-)



Same thing for profile pages, which give less information and less
links than Keith's service (dealing with events) the corresponding
dbtune service (dealing with foaf:knows and last scrobbles).

  
Yes, but note, how our works, you go to the page via 
http://ode.openlinksw.com and you simply view Page Description. We 
start by describing the document which unveils what the document is about.


I noticed that HTML5 [1] has quite an understanding (to my pleasant 
surprise) of this aspect of matters that has generally been overlooked 
in the RDF realm.


Links:

1. 
http://www.whatwg.org/specs/web-apps/current-work/multipage/microdata.html#microdata



--


Regards,

Kingsley Idehen   Weblog: http://www.openlinksw.com/blog/~kidehen
President  CEO 
OpenLink Software Web: http://www.openlinksw.com