W3C home > Mailing lists > Public > public-media-annotation@w3.org > March 2009

Re: Results of Questionnaire Survey vs. iTunes and QT

From: Tobias Bürger <tobias.buerger@sti2.at>
Date: Tue, 17 Mar 2009 23:00:23 +0100
Message-ID: <49C01D77.1090604@sti2.at>
To: "Evain, Jean-Pierre" <evain@ebu.ch>
CC: Silvia Pfeiffer <silviapfeiffer1@gmail.com>, Joakim Söderberg <joakim.soderberg@ericsson.com>, public-media-annotation@w3.org, Daniel Park <soohongp@gmail.com>
Jean-Pierre makes an important point in his mail, which is the
popularity of a format measured according to its frequency of use in
real world scenarios.

I guess he has a very good knowledge about video metadata formats
currently used and has the needed insights to know about how these
standards are really used. So I fully trust and agree to his statements
about video formats.

Personnaly I voted for formats which are prominently used also for still
images based on their (my subjective) perceived popularity. However I do
not have real numbers in front of me that proof this perceived
popularity. I fully agree that the real usage of the formats should be
another decision criteria for choosing the ones we work on. So has
anyone got real numbers about the use of the different fornats? If not:
how can we get them?

This brings me to another point: Jean-Pierre argues in favor of metadata
formats for video. I guess we should not forgot about the ones focusing
on images (e.g. EXIF, DIG35) or audio (e.g. ID3, QT, iTunes) in which
some participants have a strong interest in. At least unless we do not
discard media types from the scope of our work.

Another point I agree on that METS should be one of the first formats to
discard. METS is a container format and most notably makes use of other
formats to describe technical, descriptive or administrative metadata.
So METS as such does not provide much to map to :-)

Best regards,


Evain, Jean-Pierre schrieb:
> I don't know if we'll have to repeat the process until everyone is happy but
> as far as I am concerned I looked at what metadata is going to be available
> on the Internet for Internet TV  and Radio (and also possibly in relation to
> broadcast through content search engines) for standard non-professional
> non-academic users, which is the reason why I didn't push neither EBU's
> P-META nor SMPTE RP210 or DMS-1. 
> On the other TV-Anytime is related to so many specifications (DVB, ATIS,
> DLNA, Open TV Forum, etc.) that I shouldn't (and apparently don't) need to
> argue in favour of it, although it is 'complex' in comparison to others, By
> the way, what is the definition of complexity here: [one can't read and
> understand more than e.g. 15 elements (and too bad for the well known
> drawbacks)].
> EBUCore is going to be the base of the EUScreen project, which plans to be
> the TV and Radio archive portal of Europeana. Then, any value?
> MPEG-7. Our use of it is for the time being going to be restricted to new
> metadata extraction tools in the production. This may be used later to make
> search on 'parts' of content, but we need to show that we can produce this
> metadata at reasonable cost and then also publish it in an affordable
> manner. This means certainly not a priority one. And even if I am following
> what's  happening in mpeg's MXM project, my intention is also to see to
> which extent other metadata formats can be adapted to it and not only
> mpeg-7. Certainly, mpeg-7 doesn't at all address the description of services
> and content is not hanging in the air.
> Complexity is less an issue than looking for what described content will be
> around and what content will users want to access. I am more concerned about
> FRBR and METS as I am wondering if we are all speaking of the same thing.
> Jean-Pierre
> -----Original Message-----
> From: public-media-annotation-request@w3.org
> [mailto:public-media-annotation-request@w3.org] On Behalf Of Silvia Pfeiffer
> Sent: mardi, 17. mars 2009 22:13
> To: Joakim Söderberg
> Cc: public-media-annotation@w3.org; Daniel Park
> Subject: Re: Results of Questionnaire Survey vs. iTunes and QT
> I don't think the group has decided yet whether a low value in
> question 3 is good or bad.
> When I filled in the questionnaire, I regarded complexity as a bad thing.
> Therefore, a low number in question 3 would mean (at least for some of
> us) that it is less relevant to be examined.
> We have multiple alternate ways forward from here:
> 1. have another survey to determine whether complexity in a format is
> a preferrable
> 2. have another survey that does not ask 3 questions, but only one
> about which format the group wants to work on - that can then include
> iTunesRSS and whatever else was overseen the first time
> 3. run with the current result, accepting its shortcomings
> In my personal opinion, I don't think we need to be afraid to
> undertake this survey again. We have identified shortcomings of the
> previous approach and should address them. We're not such a large
> group that re-running the survey implies a major nuisance.
> Just my 2c. :-)
> Cheers,
> Silvia.
> 2009/3/18 Joakim Söderberg <joakim.soderberg@ericsson.com>:
>> Dear all,
>> The results from the format survey can be summarized as follows (the
> lowest sum wins):
>> Media RSS 2+1+3= 6
>> EXIF    1+5+1=  7
>> Youtube 4+2+4=  10
>> TV-Anytime 5+7+3=15
>> ID3 frames 5+4+7=16
>> MPEG-7 =8+6+2=  16
>> EBUCore 7+3+8=  19
>> This indicates the groups' average level of interest for each format.
>> Unfortunately iTunes ant QT was not added to the survey, since it appeared
> later in the mapping table. The chairs of the group regret this and it was
> recorded at today's telecom that there is an interest in the group for those
> formats as well (see minutes to appear at:
> http://www.w3.org/2009/03/17-mediaann-minutes.html). These formats are
> included in the table for review:
> http://www.w3.org/2008/WebVideo/Annotations/wiki/Format_mapping_review#Refer
> ence_Material
>> The purpose of this activity was to achieve a consensus on which formats
> to put more attention to during the review and editing process. Hopefully
> this collaborative work is of a generic character and can be reused for the
> other more complex or specific formats.
>> Regards
>> Joakim

Dipl.-Inf. Univ. Tobias Bürger

STI Innsbruck
University of Innsbruck, Austria

Received on Tuesday, 17 March 2009 21:59:12 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:17:33 UTC