Re: [MMSEM] Music Use Case & FOAF

Dear all,

Raphaël Troncy wrote:
> Dear Oscar,
>
> You did an amazing job, thank you very much!
>
> About the FOAF profile automatic extension (based on your music interest), I
> find it interesting. Do you plan to include this in the use case? 
Well, yes of course! That's one of the main reasons I did that (else I 
wouldn't implement all that code!)
> I think you should since it is just showing the new possibilities when combining
> different vocabularies ...
Indeed. It will be a part of the use case, where based on this combined 
information, it will show how to:
- get notifications of upcoming events (i.e concerts), based on the 
foaf:based_near [Already explained in the Wiki]
- generate automatic playlists based on the seed artists found on the 
user's FOAF profile and content-based audio similarities (a demo will be 
available soon, hopefully, via the /facet+music webapp). [Already 
mentioned in the Wiki, more text to fill in, though].
> About the ID3 conversion, I think that convert the data into the Music
> Ontology is an excellent choice! We should definitively try to give feedback
> to Frédérick Giasson and Yves Raimond about the practical use of this
> ontology. 
I already sent it to the Music Ontology (google groups) mailing list. So 
they were aware of that.
> I have several questions:
>     - One can give to the service a URL (with the title, author, what else?)
> to do the conversion. 
Right now only title and artist. For the moment, I think it's enough. 
More options could be the Musicbrainz/MusicDNS PUID [1].
> What is the syntax of this URL? 
Here's an example:
http://foafing-the-music.iua.upf.edu/RDFize/track?artist=u2&title=the%20fly
> Could you detail the various paramaters you can put in the URL ? All ID3 (common) tags?
Nope, only title and artist at the moment.
>     - One can alternatively give to the service a URL with a the location of
> an MP3 file. I guess that you then upload the MP3 on your server, to analyse
> it, extract the ID3 tags, and then delete the MP3 to not have problems :-)
> Correct?
Exactly, you're right.
>     - Why do you say a "possible" mapping ? Do you think we could hand up
> with different modeling choices? Could you summarize the alternatives ?
I meant "possible", in the sense that more metadata could be filled in 
(see MO examples at [2], specially level 3).
BTW, I think the conversion is quite straightforward and simple.
>
> And I'm impatient to read the summary in the Music Use Case :-)
It's on the way. I hope next week we will finish writing all the things 
we've been working on.

Thanks for all the comments!

All the best,

Oscar
[1] http://musicbrainz.org/doc/PUID
[2] http://pingthesemanticweb.com/ontology/mo/#sec-example

Received on Wednesday, 28 March 2007 08:47:41 UTC