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

Re: Updated ontology doc within Seoul F2F meeting

From: Veronique Malaise <vmalaise@few.vu.nl>
Date: Wed, 24 Feb 2010 15:47:56 +0100
Cc: 이원석 <wslee@etri.re.kr>, "public-media-annotation@w3.org" <public-media-annotation@w3.org>
Message-Id: <775501EE-C224-41B3-A2A7-510CEA43D3BF@few.vu.nl>
To: "Bailer, Werner" <werner.bailer@joanneum.at>
Dear all,

On Feb 24, 2010, at 3:25 PM, Bailer, Werner wrote:

> Dear Wonsuk, all,
>
> Thanks for the new draft, I think it is a major improvement.
>
> However, I have a few comments:
>
> - in the abstract it says we define semantics-preserving: the  
> introduction states correctly that this cannot always be achieved   
> so it should be removed from the abstract

I changed that from the abstract... maybe there is a hidden part of  
the XML document that still contains this text and that keeps on  
rewriting the latest version of the abstract?

> - section 3.1: shouldnt we list for completeness also string, float  
> and integer and refer to definitions we are using

yes, I was waiting for the list of the types to add them :)

> - section 3.2: we are currently not using any complex type  
> definitions in the table, esp. not the person example given

right, so better simplify it all and simply list the types we use,  
without making a "simple/complex" distinction

> - the editorial note in 4.1.2 should reference annex A
> - ma:location: Im not an expert in this, but afaik there are  
> different coordinate systems for geolocation, so we should add a  
> string for the coordinate system to the log/lat/alt triple

and there are also geo ontologies: URI should be a valid value; the  
same goes for other properties, we should not restrict the value to  
one specific vocabulary, but rather state that URI' and values from  
constrained vocabularies are preferred.
One question about the types: have not checked yet, but do some  
properties still have a pair as value? How "processable" is such a  
format? Is it not too much a burden and shoul dit not be replaced by  
simple value, with the help of subproperties? Maybe it's the case now,  
will check. Another point: about frameSize: having two numbers here  
too is a "hard to process" option (for mapping for example): how about  
having height/width as subproperties, at least?

> - 4.2.1.1 ma:right should be ma:copyright

I changed the ma:copyright to ma:right because of the comments of the  
PLING people, I thought they were not satisfied with copyright, that  
they wanted license and rights, but I may be mistaken

> - Annex A: I think discovery of tracks can be dropped, this is  
> resolved by the redefinition of ma:fragment
> - Annex A: I think including the references we have in the Wiki for  
> each of the issues could be useful
> - Annex C: Ive asked that before: are really all references  
> normative? E.g. MPEG-21 that we consider not in scope
> - typo: vocbulary in the introduction
> - typo: cunjuntion -> conjunction in 3.2
>
> Best regards,
> Werner
>
> From: public-media-annotation-request@w3.org [mailto:public-media-annotation-request@w3.org 
> ] On Behalf Of ???
> Sent: Mittwoch, 24. Februar 2010 09:36
> To: public-media-annotation@w3.org
> Subject: Updated ontology doc within Seoul F2F meeting
>
> below is the updated ontology doc within Seoul F2F meeting.
> Please review and if you have any comment, let me know.
>
> http://dev.w3.org/2008/video/mediaann/mediaont-1.0/mediaont-1.0.html
>
> best regards,
> Wonsuk.
>
Received on Wednesday, 24 February 2010 14:49:32 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 24 February 2010 14:49:34 GMT