- From: Chris Poppe <Chris.Poppe@ugent.be>
- Date: Tue, 19 Oct 2010 15:04:59 +0200
- To: "'Doug Schepers'" <schepers@w3.org>, 'Joakim Söderberg' <joakim.soderberg@ericsson.com>
- Cc: "'Daniel Park'" <soohongp@gmail.com>, '이원석' <wslee@etri.re.kr>, <tmichel@w3.org>, <public-media-annotation@w3.org>
Hi Doug,
For me the phone conf can be hold on Thursday (21st of October) or Friday (22nd) 15.00 (European time).
For your scenario, please see my comments ([CP]) on your example below, based on the latest version of the API document [0].
Note that I have added the public mailing list in cc since I think this is interesting for other people as well.
[0] http://dev.w3.org/2008/video/mediaann/mediaont-api-1.0/mediaont-api-1.0.html
Kind regards,
Chris
Ghent University - IBBT
Faculty of Engineering
Department of Electronics and Information Systems (ELIS)
Multimedia Lab Gaston Crommenlaan 8 bus 201
B-9050 Ledeberg-Ghent
Belgium
t: +32 9 33 14959
f: +32 9 33 14896
t secr: +32 9 33 14911
e: chris.poppe@ugent.be
URL: http://multimedialab.elis.ugent.be
-----Original Message-----
From: Doug Schepers [mailto:schepers@w3.org]
Sent: Monday, October 18, 2010 11:25 AM
To: Joakim Söderberg
Cc: Chris.Poppe@UGent.be; 'Daniel Park'; ???; tmichel@w3.org
Subject: Re: [AGENDA] Media Annotations WG Teleconf - 2010-10-12
Hi, Joakim-
Joakim S�derberg wrote (on 10/12/10 8:01 AM):
> The planned Telecon today at 15.00 (European time) is cancelled. Chris
> and Wonsuk have proposed Thursday or Friday at 15.00, can you make it?
>
> May I also take the opportunity to remind you of the scenarios that you
> were to work out?
Last week turned out to be busier than I anticipated. Perhaps we could
meet later this week instead, and I will try to have the scenarios done
by then?
As a rough example, take the following file:
An MP3 of a music track by (fictional) band "The Foos", with the title
"Separated" off the album "Specks", with multiple annotations:
1) an ID3 of the band name, song title, and album title, where the song
title includes the band name ("The Foos - Separated"), dated 2009-12-02,
by person A
2) an XMP of the band name, song title, and album title, where the song
title does not include the band name, dated 2010-06-14, by person B
3) some 3rd format (or ID3 or XMP, you decide) with the band name, song
title, and a URL to an HTML or text page with the lyrics to the song,
undated, by person C
How would you perform the following tasks:
1) Get only the annotations in XMP format?
[CP] I solved this task in three ways, one retrieves all titles in XMP, another retrieves all properties in XMP, the third one retrieves all metadata in XMP.
//first create an object implementing the MediaResource interface
mediaResource = new MediaResource("http://someID");
//mediaResource provides access to the metadata of the media resource identified by the id
//get all titles in XMP
status = mediaResource.getMediaProperty("title",titleProperty,"","XMP");
/*
Resulting in:
status = "OK";
titleProperty[0].value = "Separated"
titleProperty[0].type = "Song title"
titleProperty[0].language = "en-us"
titleProperty[0].sourceFormat = "XMP"
titleProperty[1].value = "Specks"
titleProperty[1].type = "Album title"
titleProperty[1].language = "en-us"
titleProperty[1].sourceFormat = "XMP"
*/
//get all properties of XMP
//first ask which properties hold values, this is returned in the properties array
status = mediaResource.getPropertyNamesHavingValues(properties,"XMP");
//next go over the properties array and request for the values if wanted
for(j = 0; j<properties.length; j++)
{
status= mediaResource.getMediaProperty(properties[j],tempProp);
}
//get all metadata of XMP
status = getOriginalMetadata("XMP", metadataHolder);
/*
Resulting in:
status = "OK";
metadataHolder[0] = "...<dc:title>Separated</dc:title><xmpDM:album>Specks</xmpDM:album><xmpDM:artist>The Foos</xmpDM:artist>..."
*/
2) Get only the annotations from person A?
[CP] We can't, there is no provenance information on the metadata-level.
3) Find all titles?
[CP]
//get all titles
status = mediaResource.getMediaProperty("title",titleProperty);
/*
Resulting in:
status = "OK";
titleProperty[0].value = "The Foos - Separated"
titleProperty[0].type = "Song title"
titleProperty[0].language = "en-us"
titleProperty[0].sourceFormat = "ID3"
titleProperty[1].value = "Specks"
titleProperty[1].type = "Album title"
titleProperty[1].language = "en-us"
titleProperty[1].sourceFormat = "ID3"
titleProperty[2].value = "Separated"
titleProperty[2].type = "Song title"
titleProperty[2].language = "en-us"
titleProperty[2].sourceFormat = "XMP"
titleProperty[3].value = "Specks"
titleProperty[3].type = "Album title"
titleProperty[3].language = "en-us"
titleProperty[3].sourceFormat = "XMP"
*/
4) Find only the song title(s)?
[CP]
//get all song titles in XMP
status = mediaResource.getMediaProperty("title",titleProperty, "","","Song title");
/*
Resulting in:
status = "OK";
titleProperty[0].value = "The Foos - Separated"
titleProperty[0].type = "Song title"
titleProperty[0].language = "en-us"
titleProperty[0].sourceFormat = "ID3"
titleProperty[1].value = "Separated"
titleProperty[1].type = "Song title"
titleProperty[1].language = "en-us"
titleProperty[1].sourceFormat = "XMP"
*/
5) Find the most recent (latest annotation) song title?
[CP] We can't, there is no date information on the metadata-level.
6) Retrieve external lyric metadata? (Consider if the metadata were
something other than the lyrics, like the song title.)
[CP] The api disregards where the metadata is stored, an implementation of this API should try to find relevant metadata sources.
Regards-
-Doug Schepers
W3C Team Contact, SVG and WebApps WGs
Received on Tuesday, 19 October 2010 13:06:42 UTC