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

RE : RE : [mawg] action-249: Ontology Rev. 7 available

From: Evain, Jean-Pierre <evain@ebu.ch>
Date: Tue, 14 Sep 2010 19:52:38 +0200
To: "Bailer, Werner" <werner.bailer@joanneum.at>, Tobias Bürger <tobias.buerger@salzburgresearch.at>
CC: "public-media-annotation@w3.org" <public-media-annotation@w3.org>
Message-ID: <7D1656F54141C042A1B2556AE5237D60010CEEB045EC@GVAMAIL.gva.ebu.ch>
Dear Werner,

I'll let Tobias respond but I believe you are missing the point entirely :-(

In fact the ma-ont ontoloy  like DC (and subsequently e.g. EBUCore or PBCore) have identified Creator and Publisher or Actor as sufficiently important to coexist with a more generic Contributor term. Why because it was thought to be important for the user queries. 

We are in a paradigm different from xml schema in which the above mentioned schemas have been defined.

In order to benefit from richer queries on contributor the proposed approach is much more powerful, reestablish a logical relationship between contributors, creators, publishers and actor while it still fulfils the requirement for searching these sub classes.

This is what I plan to do for EBUCore too.

If some relations are missing we can still add them. I have seen the proposal from veronique to add first publsihed / broadcast date, which I can agree with as we have always had this in EBUCore ;-)

I believe this work on the rdf ontology is what I have always this group should do. This will eventually be much more powerful than any java API. Let's not undermine this. It should rather be the list of attributes and properties that should be amended to match the need of a good ontology.

Is there any other reason why you are so reluctant? Is it in conflict with some of your mpeg-7 developments? Of course, this could not quite justify changes, could it?

Best regards,

De : public-media-annotation-request@w3.org [public-media-annotation-request@w3.org] de la part de Bailer, Werner [werner.bailer@joanneum.at]
Date d'envoi : mardi, 14. septembre 2010 18:12
À : Tobias Bürger
Cc : public-media-annotation@w3.org
Objet : AW: RE : [mawg] action-249: Ontology Rev. 7 available

Dear Tobias,

thanks for your clarifications. I agree that forcing the ontology to be flat list misses the goal, and there is the need to introduce concepts and properties for modelling properties. However, I would expect that the differences between the description of the properties and in the ontology document and the ontology are only technical.

- creator/contributor issue: The ontology defines a subclass relation between the two, while in my assumption the ontology document defines them as disjoint (at least as I understand the document). We have not defined subtypes of our properties, but use attributes to qualify them. If there is a good reason for having this subclass relation for creator and contributor, then we should say so in the ontology document, and - to keep a consistent approach - have contributor, with creator being one of the values for contributor.type.

- actor: As I mentioned above, we have refrained from defining subtypes. So why make this particular exception? For me this could serve as an example of how to extend the core ontology, but I do not think we should include it in an ontology that we consider normative.

Best regards,

Von: public-media-annotation-request@w3.org [public-media-annotation-request@w3.org] im Auftrag von Tobias Bürger [tobias.buerger@salzburgresearch.at]
Gesendet: Dienstag, 14. September 2010 12:06
An: public-media-annotation@w3.org
Cc: public-media-annotation@w3.org
Betreff: Re: RE : [mawg] action-249: Ontology Rev. 7 available

  Dear all,

let me step into this discussion.

First of all, I guess the fact that we decided that the Media Ontology
as it is defined in our document is a flat list of properties should not
restrict ourselves in building an ontology (using an ontology definition
language) that models the universe captured by the Media Ontology more
precisely (meaning with more precise semantics). Having said that, I am
strongly in favor of defining creator to be a subclass of contributor if
we think the semantics of this relationship hold. For me a creator of a
media resource is always a contributor and thus this relationship holds
in my opinion. The same is true for the other roles (i.e. actor and
publisher also are some type of contributor).

There was also the issue about Actor. If I recall correctly, then we
said in one meeting that we would like to have Actor as a class (F2F in
Vienna). But if the group thinks we should mark it as an extension, we
can do that.

With respect to some properties/concepts we introduced in addition to
the core MAWG properties (defined in the document), they do in most
cases define the relations between concepts defined in our ontology more
clearly and make them explicit; such as isMadeOf to be able to relate
tracks to MediaResources. Also we introduced Agents (Persons or
Organisations) to be able to express who can be creator or contributor
of a resource. I definitely would keep this. If the group decides to
remove all of this, then we end up with a flat list of properties again
(without defining any precise semantics) and this is not what we want at
this point, I'd say.

Best regards,


Am 08.09.2010 17:13, schrieb Evain, Jean-Pierre:
> Dear Werner,
> For the choice of the classes, it was the MAWG choice to have actor. And to be consistent actor is a class and not a property.
> I wouldn't undermine the MAWG ontology e.g. because mpeg did some wrong ontological assumptions. Look at the questiosn I asked and must be asked to decide if classes are sub-classes and o what. The current classification is natural.
> Definitely I beleiev being able to factor queries is clearly and advantage. this is also one of the reasons why we introduced the notion of agent.
> Best regards,
> Jean-pierre
> ________________________________________
> De : Bailer, Werner [werner.bailer@joanneum.at]
> Date d'envoi : mercredi, 8. septembre 2010 16:26
> À : Evain, Jean-Pierre; Tobias Bürger
> Cc : public-media-annotation@w3.org
> Objet : AW: [mawg] action-249: Ontology Rev. 7 available
> Dear Jean-Pierre,
> then main difference is that there is a class Actor in the ontology, but no corresponding property in our table (meaning property in the sense of  "core property" in our specification, which does not necessarily mean that this is modelled as a property in the ontology). Actor is one of the many possible subtypes of contributor, which we decided not to exhaustively list in our spec.
> Concerning the question of contributor and creator, I think it should be as much aligned with our list of core properties as possible, as well as the formats we support in the mapping table. And as far as I see from the mapping tables, some formats have them separate, one uses Creator as the superclass, some make no distinction. So whatever we choose, it will not be perfectly aligned with any of the formats. But I do not see a reason for defining creator as a subclass of contributor, rather for keeping them separate.
> Best regards,
> Werner
> ________________________________________
> Von: Evain, Jean-Pierre [evain@ebu.ch]
> Gesendet: Mittwoch, 08. September 2010 16:02
> An: Bailer, Werner; Tobias Bürger
> Cc: public-media-annotation@w3.org
> Betreff: RE: [mawg] action-249: Ontology Rev. 7 available
> By the way I fail to see how the current owl file is different from our property definitions (this without saying that I feel some vagueness in the notions of class vs. Property in your comment - with all due respect Werner ;-)
> JP
> -----Original Message-----
> From: public-media-annotation-request@w3.org [mailto:public-media-annotation-request@w3.org] On Behalf Of Bailer, Werner
> Sent: mercredi, 8. septembre 2010 15:41
> To: Tobias Bürger
> Cc: public-media-annotation@w3.org
> Subject: AW: [mawg] action-249: Ontology Rev. 7 available
> Dear Tobias,
> thanks for this draft. I have the following comments:
> - As I posted to the list earlier, I wonder why Creator is a subclass of Contributor. In our list of properties they are disjoint, as in some other formats (e.g. EBUCore), in MPEG-7 it's even the other way round (Creator being the superclass): I'm in favour of keeping them separate, as I think it's confusing to model that differently than in our properties definition.
> - We should be careful to align the ontology with our list of properties, e.g. Actor is useful as an example, but not defined in our list of properties, thus it should not be included in the official ontology that we publish.
> Best regards,
> Werner
> ________________________________________
> Von: public-media-annotation-request@w3.org [public-media-annotation-request@w3.org] im Auftrag von Tobias Bürger [tobias.buerger@salzburgresearch.at]
> Gesendet: Mittwoch, 08. September 2010 14:28
> An: public-media-annotation@w3.org
> Betreff: [mawg] action-249: Ontology Rev. 7 available
>    Dear all,
> after some discussions in the previous week, we made an update to the
> ontology which you can find at
> http://www.salzburgresearch.at/~tbuerger/ma-ont-rev7.owl
> What we changed:
> We changed the things to which Pierre-Antoine referred to in his
> comments 1 and 3 in
> http://lists.w3.org/Archives/Public/public-media-annotation/2010Aug/0027.html,
> that is, we fixed the properties with multiple domain declarations
> (e.g., title and locator) and we adapted the Contributor subclass hierarchy.
> Furthermore we added a new property isMadeOf to express that a
> MediaResource can be made of Audio - and/or VideoTracks.
> I would like to note that we did not change the modelling wrt. to the
> restricted use of some properties (.e.g duration does not apply to
> Images) as both discussed options have their drawbacks in the one or
> other direction (e.g. when introducing new subclasses of MediaResource
> or when adding new properties whose use shall be restricted somehow).
> The same arguments as the ones from Pierre-Antoine against the current
> solutions can be made with his proposal because what, for instance,
> ma:Image rdfs:subClassOf ma:MediaResource, [
>       a owl:Restriction ;
>       owl:onProperty ma:duration ;
>       owl:cardinality 0
>     ]
> tells us is, that Images are MediaResources which do not have a duration
> property. This gives imho also gives a wrong impression.
> I had a discussion internally in my group and without telling them what
> options are being discussed at the moment, most of them favored the
> modelling which we currently have in our version.
> Both aspects lead me to the conclusion that I keep this part of our
> ontology as it is for now. Unless new arguments or majority votes
> against our current modelling pop up :-)
> Any feedback is again highly welcome!
> @Thierry: If there are no objections you might publish this draft in the
> W3C web space.
> Best regards,
> Tobias
> --
> ================================================================
> Dr. Tobias Bürger         Knowledge and Media Technologies Group
> Salzburg Research                           FON +43.662.2288-415
> Forschungsgesellschaft                      FAX +43.662.2288-222
> Jakob-Haringer-Straße 5/III   tobias.buerger@salzburgresearch.at
> A-5020 Salzburg | AUSTRIA         http://www.salzburgresearch.at

Dr. Tobias Bürger         Knowledge and Media Technologies Group
Salzburg Research                           FON +43.662.2288-415
Forschungsgesellschaft                      FAX +43.662.2288-222
Jakob-Haringer-Straße 5/III   tobias.buerger@salzburgresearch.at
A-5020 Salzburg | AUSTRIA         http://www.salzburgresearch.at
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

Received on Tuesday, 14 September 2010 17:53:25 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 16:24:42 UTC