- From: Raphaël Troncy <Raphael.Troncy@cwi.nl>
- Date: Thu, 26 Oct 2006 11:28:06 +0200
- To: MMSem-XG Public List <public-xg-mmsem@w3.org>, Stamatia Dasiopoulou <dasiop@iti.gr>, Vassilis Tzouvaras <tzouvaras@image.ntua.gr>, Michael Hausenblas <michael.hausenblas@joanneum.at>, Suzanne Little <Suzanne.Little@isti.cnr.it>
Dear Stamatia, Michael and Vassilis, Thank you for your work on this use case. I think the work performed these last days is important and I see this use case ready to be reviewed now. I have some comments: - In your introduction, you mention 2 examples illustrating the internal interoperability problems of MPEG-7. The first one is pretty clear and nicely examplified. I will tend to disagree on your second one. If I agree that MPEG-7 does not provide any formal semantics, and therefore no semantics to the relation between a whole (multimedia content) and its constituent parts, I have found your example misleading. In MPEG-7, everythning is a Segment. Therefore, a StillImage is a Segment, and its constituent are also Segment. If you query for a Segment depicting Zidanne, you will get BOTH the whole image or the individual segments annotated with him, with a simple XPath query. Now, if you would like to mention you need some part-whole reasoning, it is a different issue and you need to rephrase your example. More precisely, in your case, if the Segment "only" is annotated with "Zidane" but not the whole image, would you like to return the whole image to the query ? - I have slightly edited your text putting a new section when you described the existing MPEG-7 ontologies. I found VERY interesting your attempt to use the MPEG-7/ABC ontology proposed by Jane Hunter ... We might all need here to have the wise comments of Suzanne Little who worked with Jane. Suzanne, you will be most likely nominate to review this use case :-) - In the new section 2.2, could you give also examples showing where does it break between the Tsinaraki formalization of the SemanticsDS and a possible coupling of the Hunter Ontology with some domain specific ontology ? Could this problem not be simply solved with some owl:sameAs statements ? -The Section 3 is promising and gives interesting research directions ... How much you plan to elaborate on each ? Best regards. Raphaël -- Raphaël Troncy CWI (Centre for Mathematics and Computer Science), Kruislaan 413, 1098 SJ Amsterdam, The Netherlands e-mail: raphael.troncy@cwi.nl & raphael.troncy@gmail.com Tel: +31 (0)20 - 592 4093 Fax: +31 (0)20 - 592 4312 Web: http://www.cwi.nl/~troncy/
Received on Thursday, 26 October 2006 09:45:43 UTC