W3C home > Mailing lists > Public > public-media-annotation@w3.org > December 2010

Re: RE : RE : Next iteration of the RDF ontology

From: Pierre-Antoine Champin <pierre-antoine.champin@liris.cnrs.fr>
Date: Mon, 06 Dec 2010 10:39:12 +0100
Message-ID: <4CFCAF40.40106@liris.cnrs.fr>
To: "Evain, Jean-Pierre" <evain@ebu.ch>
CC: "'tobias@tobiasbuerger.com'" <tobias@tobiasbuerger.com>, "mcsuarez@fi.upm.es" <mcsuarez@fi.upm.es>, "public-media-annotation@w3.org" <public-media-annotation@w3.org>
On 12/06/2010 10:07 AM, Evain, Jean-Pierre wrote:
> But an image could be a fragment covering the whole picture.

Ok; but then what you call "picture" would not be an instance of class 
Image?? Not sure I'm following you...

> This is
> the logic followed by MFWG whose fragment URI can access a media
> resource or a fragment.

Here again, I am not sure I'm following you, but you are raising an 
interesting point. Consider the three following resources:

   :a = http://example.org/photo.png .
   :b = http://example.org/photo.png?xywh=percent:0,0,100,100 .
   :c = http://example.org/photo.png#xywh=percent:0,0,100,100 .

What would be their respective types in your vision of the ontology? 
What would be their ma:hasFragment relations ? Would any of them be 
owl:sameAs others?

This is how I see things for the moment:

   :a a ma:Image ; ma:hasFragment :b, :c .
   :b a ma:Image, ma:MediaFragment .
   :c a ma:Image, ma:MediaFragment .

Note that IMHO none of them is owl:sameAs another one (but this is just 
precaution).

Note also that I make no difference between :b and :c, which is maybe a 
shame; on the other hand, my understanding of 'Media Fragments URI 1.0' 
is that Media Fragment subsumes URI fragment and URI query, so 
ontologically it does not seem wrong.

   pa



>
> JP
>
> -----Original Message----- From: Pierre-Antoine Champin
> [mailto:pierre-antoine.champin@liris.cnrs.fr] Sent: lundi, 6.
> décembre 2010 10:05 To: Evain, Jean-Pierre Cc:
> 'tobias@tobiasbuerger.com'; mcsuarez@fi.upm.es;
> public-media-annotation@w3.org Subject: Re: RE : RE : Next iteration
> of the RDF ontology
>
> On 12/06/2010 08:01 AM, Evain, Jean-Pierre wrote:
>> Then the logic would be to move Image as a sub-class of fragment??
>
> I don't think so: *some* images are not a fragment of anything, but
> "top level" media resources...
>
> The ontology should allow for a resource to be both an Image and a
> MediaFragment, but this is already the case, since those classes are
> not disjoint. I think it is enough.
>
> pa
>
>>
>> JP
>>
>> ________________________________________ De : Pierre-Antoine
>> Champin [pierre-antoine.champin@liris.cnrs.fr] Date d'envoi :
>> dimanche, 5. décembre 2010 22:35 À : Evain, Jean-Pierre Cc :
>> 'tobias@tobiasbuerger.com'; mcsuarez@fi.upm.es;
>> public-media-annotation@w3.org Objet : Re: RE : Next iteration of
>> the RDF ontology
>>
>> On 12/03/2010 01:18 PM, Evain, Jean-Pierre wrote:
>>> I was thinking of this but I am not sure that we have any
>>> mechanism to point to a fragment / region within a picture ->
>>> at least not covered by the ontology and I am not even sure about
>>> what MFWG has done, which would allow their URI to point to one.
>>
>> Media Fragment URIs allows for rectangular spatial fragments:
>> http://www.w3.org/TR/media-frags/#naming-space
>>
>> pa
>>
>>>
>>> Regards,
>>>
>>> Jean-Pierre
>>>
>>>
>>> -----Original Message----- From: Pierre-Antoine Champin
>>> [mailto:pierre-antoine.champin@liris.cnrs.fr] Sent: vendredi, 3.
>>> décembre 2010 12:29 To: Evain, Jean-Pierre Cc:
>>> 'tobias@tobiasbuerger.com'; mcsuarez@fi.upm.es;
>>> public-media-annotation@w3.org Subject: Re: RE : Next iteration
>>> of the RDF ontology
>>>
>>> On 12/03/2010 09:51 AM, Evain, Jean-Pierre wrote:
>>>> Dear Mari-Carmen,
>>>>
>>>> Based on the latest version (thanks Tobias ;-), we could
>>>> effectively be more restrictive and say that MediaFragment
>>>> isFragmentOf (MediaResource and not Image).
>>>
>>> ehr... an Image can have fragments, namely spatial fragments.
>>>
>>> In general, to respond Mari's comment about constraining
>>> hasFragment is a two side coins... By constraining, we may indeed
>>> detect some inconsistencies... On the other hand, we might limit
>>> the use of the ontology in situations that we do not envision
>>> right now.
>>>
>>> So I would be in favor of leaving the domain and range as is. A
>>> specific application is of course free to put additional
>>> constraints to fulfill its needs.
>>>
>>> This is a personal opinion though; not necessarily the one of
>>> the RDF Taskforce or the WG...
>>>
>>> pa
>>>
>>>
>>>>
>>>> If I have covered most of your questions in my two mails then
>>>> I’ll work on a version 26. Waiting for confirmation.
>>>>
>>>> Best regards,
>>>>
>>>> Jean-Pierre
>>>>
>>>> *From:*tobias.buerger@gmail.com
>>>> [mailto:tobias.buerger@gmail.com] *On Behalf Of *Tobias Bürger
>>>> *Sent:* vendredi, 3. décembre 2010 08:33 *To:* Evain,
>>>> Jean-Pierre *Cc:* mcsuarez@fi.upm.es; Pierre-Antoine Champin;
>>>> public-media-annotation@w3.org *Subject:* Re: RE : Next
>>>> iteration of the RDF ontology
>>>>
>>>> Dear Mari-Carmen,
>>>>
>>>> thanks also from my side for the feedback and thanks to
>>>> Jean-Pierre for answering your questions!
>>>>
>>>> What I wanted to add is, that you, Mari-Carmen, looked at an
>>>> old version of the ontology. The most recent version was sent
>>>> around with this mail:
>>>> http://lists.w3.org/Archives/Public/public-media-annotation/2010Nov/0130.html
>>>>
>>>>
>>>>
>>
>>>>
Best regards,
>>>>
>>>> Tobias
>>>>
>>>> 2010/12/2 Evain,
>>>> Jean-Pierre<evain@ebu.ch<mailto:evain@ebu.ch>>
>>>>
>>>> Hello Mari-Carmen,
>>>>
>>>> Thanks for the feedback.
>>>>
>>>> I'll first try to summarise what the intention was and then
>>>> we'll come back to your specific points.
>>>>
>>>> The idea of the current class model is:
>>>>
>>>> A MediaResource can be one or more images and /or one or more
>>>> AV MediaFragment.
>>>>
>>>> By definition, in the model, an AV MediaResource is made of at
>>>> least one MediaFragment.
>>>>
>>>> A MediaFragment is the equivalent of a segment or in some
>>>> standards like NewsML-g2 or EBUCore, a part.
>>>>
>>>> A MediaFragment is composed of one or more media components
>>>> organised in tracks (separate tracks for captioning/subtitling
>>>> or signing if provided in a separate file): audio, video,
>>>> captioning/subtitling, signing. There could be other types of
>>>> tracks like a 'data' track, etc.
>>>>
>>>> Addressing some of your remarks:
>>>>
>>>> - a frame could be a MediaFragment with a duration of one frame
>>>> and if you wnat to address only the farme as a video frame then
>>>> the component is the VideoTrack. We could have segment and
>>>> frame as possible media fragments in the definition - an image
>>>> could also be a key frame - as mentioned above captioning is
>>>> the same as subtitle and this should be mentioned in the
>>>> definitions if you think it helps.
>>>>
>>>> For isFragmentOf, I'll come back to you tomorrow.
>>>>
>>>> It took me 48 hours to return from Paris making me a climatic
>>>> refugee going from airports to train stations. That's exactly
>>>> when my main PC decide to crash and doesn't let me log in. I am
>>>> working from a backup PC on which I don't have the last version
>>>> of the ontology. SHould be fine by tomorrow ;-)
>>>>
>>>> Best regards,
>>>>
>>>> Jean-Pierre
>>>>
>>>>
>>>>
>>>> ________________________________________ De : Mari Carmen
>>>> Suárez de Figueroa Baonza
>>>> [mcsuarez@fi.upm.es<mailto:mcsuarez@fi.upm.es>] Date d'envoi :
>>>> jeudi, 2. décembre 2010 17:17 À : Evain, Jean-Pierre Cc :
>>>> Pierre-Antoine Champin; public-media-annotation@w3.org
>>>> <mailto:public-media-annotation@w3.org>   Objet : Re: Next
>>>> iteration of the RDF ontology
>>>>
>>>>
>>>> Dear Jean-Pierre and all,
>>>>
>>>> I took a look to the ontology you sent on 15th November, and I
>>>> have a pair of comments (maybe you have already discussed
>>>> about them, sorry if this is the case).
>>>>
>>>> - With respect to the Track class and its subclasses
>>>> (AudioTrack, Captioning, VideoTrack), I would suggest to
>>>> complete the comments for the subclasses, because as it is know
>>>> is difficult to understand the meaning of them (for a
>>>> newcomer). In this context I have a pair of doubts: is it
>>>> AudioTrack the same as Segment? is it VideoTrack the same as
>>>> Frame? is it Captioning the same as Subtitle? If so, could you
>>>> consider to include these labels as synonyms of the existing
>>>> classes?
>>>>
>>>> - In the case of the relation called "isFragmentOf" (domain:
>>>> MediaFragment; range: MediaResource), I was wondering if it
>>>> would not be better to extend/modified the current modelling in
>>>> order to avoid possible inconsistences (such as "an image
>>>> having as a fragment a video track and an audio track").
>>>>
>>>> Thank you very much in advance. Best Regards,
>>>>
>>>> Mari Carmen.
>>>>
>>>> Evain, Jean-Pierre escribió:
>>>>> Dear all,
>>>>>
>>>>> Following the changes made during TPAC, we have been working
>>>>> with
>>>> Pierre-Antoine and Tobias to improve the ontology and the
>>>> mapping to the abstract ontology.
>>>>>
>>>>> The result of this work is attached. We will suggest a few
>>>>> changes to
>>>> the abstract ontology to improve the logic of the semantic
>>>> (date property structure) and also to improve interoperability
>>>> with the MFWG specification (improving the mediaFragment
>>>> structure).
>>>>>
>>>>> You will also notice that we are now more systematic in our
>>>>> approach
>>>> illustrated by the removal of the contributor class hierarchy
>>>> (which was there to mimic the abstract structure and help
>>>> adoption) now implemented through properties.
>>>>>
>>>>> Pierre Antoine will review the mapping table and we'll update
>>>>> the RDF
>>>> according to the decisions we make tomorrow.
>>>>>
>>>>> Cheers, JP (also on behalf on Tobias and Pierre-Antoine)
>>>>>
>>>>>
>>>>>
>>>>> -----------------------------------------
>>>>> ************************************************** This email
>>>>> and any files transmitted with it are confidential and
>>>>> intended solely for the use of the individual or entity to
>>>>> whom they are addressed. If you have received this email in
>>>>> error, please notify the system manager. This footnote also
>>>>> confirms that this email message has been swept by the
>>>>> mailgateway
>>>>> **************************************************
>>>>>
>>>>
>>>> -- ---------------------------------------------- Dr. Mari
>>>> Carmen Suárez-Figueroa Teaching Assistant
>>>>
>>>> Ontology Engineering Group (OEG)
>>>>
>>>> Departamento de Inteligencia Artificial Facultad de
>>>> Informática Universidad Politécnica de Madrid Campus de
>>>> Montegancedo, s/n Boadilla del Monte - 28660 Madrid
>>>>
>>>> Phone: (+34) 91 336 36 72 Fax: (+34) 91 352 48 19 e-mail:
>>>> mcsuarez@fi.upm.es<mailto:mcsuarez@fi.upm.es>   Office: 3205
>>>> ----------------------------------------------
>>>>
>>>>
>>>>
>>>>
>>>> -- ___________________________________ Dr. Tobias Bürger
>>>> http://www.tobiasbuerger.com
>>>>
>>>
>
Received on Monday, 6 December 2010 09:40:12 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 6 December 2010 09:40:12 GMT