W3C home > Mailing lists > Public > public-swbp-wg@w3.org > October 2005

Re: [MM] Timing of Mike & Libby's review

From: Jacco van Ossenbruggen <Jacco.van.Ossenbruggen@cwi.nl>
Date: Wed, 19 Oct 2005 10:54:40 +0200
Message-ID: <435609D0.3080104@cwi.nl>
To: "Uschold, Michael F" <michael.f.uschold@boeing.com>
CC: Libby Miller <libby@asemantics.com>, swbp <public-swbp-wg@w3.org>

Uschold, Michael F wrote:

>Yes, I will. I couldn't stop myself from having a peak at the draft as
>of Friday,
Mike, That was what I hoped for when sending you the URI :-)

> Here are a few comments:
>You make a very bold claim: 
>"The best solution to this problem covering almost all the present and
>future uses of the content is the description and archiving of each
>photo with the aid of a tool providing a semantic metadata structure
>using the Semantic Web technologies "
>It is an untested hypothesis that semantic approaches will be the best,
>it is also untested that Semantic Web languages are the best of all
>semantic approaches.
>It may be time to speculate that one day that might be true, but it is
>clearly not so today. 
>It would be truer to say:
>"The best solution to this problem covering almost all the present and
>future uses of the content is the description and archiving of each
>photo with the aid of a tool providing a metadata structure for
>organizing the photo collections."
I agree this was too bold a claim to make.  Additionally, I want no 
discussion about what good and bad solutions are in this section.  The 
role of the use case section is to illustrate  the problems in way 
people can relate to.  Later in the document we will provide (parts of) 
example solutions for each use case using the vocabularies and tools 
described.  This is also the place to discuss advantages and 
disadvantages of the solution discussed. So I've simply removed the 
entire paragraph starting from "The best solution ..." till the end of 
the use case.

>The best solution for most people today, is to use any of the myriad of
>photo organization tools out there from the simple and free ones, to the
>high end pro-oriented ones.  You don't even mention those.
Agreed. The relation between Semantic Web-based approaches and others 
will be the topic of the second deliverable. I've tried to make this 
clear in the second para of the abstract.  If this need more emphasis, 
please let me know.

>Obviously, this note does not address users of these tools, but rather
>developers.  I'm a serious photographer drowning in over 20,000 slides
>and another few thousand jpegs. I recognized 15 years ago that I really
>needed an ontology to categorize them all. I'm still waiting for tools
>to support this.
Yes, this frustration was my main motivation to setup this TF!

>I would like to see this document specifically target developers of
>photo organizing software for my own selfish reasons, I want to have
>semantically-enabled tools!
Frankly, I've been worried about the broad target audience we identify 
in the beginning of the document.  We would be extremely happy if this 
document convinces mainstream developers to start SW-enabled image 
annotation tools, but we also want to convice (professional) end-users 
that this is the way to go and they should ask for/buy SW-enabled tools, 
and convince the online/flickr/blogging community ...  Maybe this is a 
bit too ambitious.
But I'm open for tips how we can do a better job targetting developers 
of photo organizing software.

>I would love to see a new wave of photo-organization tools with semantic
>tagging and search. By far the richest one around, of the affordable
>variety is: Imatch http://www.photools.com/ which has a remarkably rich
>feature set. 
>See: http://www.urban75.org/tech/imatch.html for a [random and possibly
>dated] review.
>I don't know that such tools could be affordably created for a mass
>market, but they might work for the pro-sumer and pro markets. One
>potential benefit in the mass market is for people to share their
>ontologies using OWL mapping tools. I wish I got paid to do this stuff
>applied to photography.
Nick, do you have time to look into Imatch?  Thanks!

>"Use Case: large-scale image collections at NASA"
>This is too specific, why single out NASA, lots of organizations have
>large image collections.  Large companies have them (e.g. Boeing), so do
>the many commercial image libraries, such as Corbus, or Getty.
We want to provide concrete (thus specific) use cases with concrete  and 
specific examples.  The next version of the doc will have concrete 
images from NASA with concrete RDF/OWL annotations.  I hope that these 
specific examples will provide more insight than generalized ones, and 
that people will recognize how it relates to there own problems.  If the 
example solutions work as well in a generalized form, we will of course 
do so.

>These collections exist for different reasons, for a big company, images
>are used for a variety of reasons, such as PR, internal presentaions,
>and some may be sold externally. For an image library, the sole purpose
>is to sell images. These give rise to different 'sub-use-cases, that may
>differ in important ways.
Yes, I agree there are different use cases within the same organisation, 
but I do not think this should be reflected in the use case structure (I 
do not want to discuss Use Case :-)

I hope the final set of use cases is sufficiently representative so 
organisations recognize their different problems in different use cases.

 Thanks for the feedback,

Received on Wednesday, 19 October 2005 08:55:00 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:09:44 UTC