I'd like to offer a conjecture. Forgive me if I have
stepped into a thread without paying close enough attention
to some of the details. Recently, I delivered a paper at a
topic maps workshop in Leipzig [1] based on the notion of
"just for me", in which I was pointing out some of the
mismatches we were finding as we observed users of IRIS
[2]. IRIS is an ontology-driven semantic desktop
application, the user interface to the CALO project funded
by Darpa and being crafted under the guidance of people at
SRI. The thesis of my paper was that it is possible to wrap
an ontology with a topic map. The two, sometimes competing,
goals are to satisfy the needs of users who are not
ontologists, and, at the same time, maintain semantic
interoperability among platforms. (01)
Steven Newcomb, who has talked before this group in the
past, has, in the latest renditions of topic mapping
expressed in the new TMRM documents, pointed out that we
(topic map designers) work really hard to maintain ontology
neutrality, as much as anything, to maintain a free market
for ideas to be expressed. When ontologies are federated
through expression in a topic map, occasionally, subjects
merge, forming bonds between often disparate ontologies,
providing a means by which different world views can be
seen, discussed, and applied. (02)
"Just for me" stems from the notion that a workstation
really serves its user first; gestures expected of the user
ought to reflect those the user is comfortable with. To ask
a user to understand that a person is an instance of
something with a latin name just, it turns out, doesn't
fly. We found a use case where it is better to let the user
offer up personal names and relationships for things, not
unlike the social computing trend to move to folksonomies
and tagging at the expense of precise ontologies. I believe
that the federation of the needs of users along with the
different world views expressed in different ontologies can
be facilitated by injecting a topic map into the mix. (03)
Right now, it's just a conjecture. That might change given
that I am involved in a project to wrap ontologies with a
topic map. I'll turn the lights back on now. (04)
Jack
[1] http://www.informatik.uni-leipzig.de/~tmra05/
[2] http://www.openiris.org/ (05)
On Tue, 7 Feb 2006 15:14:15 -0500
dbedford@xxxxxxxxxxxxx wrote:
>Nicolas,
>
>This is probably a contribution I could make. I'll need
>to know more about how
>people want to access/use content to make sure that the
>profile and data model
>are on target. We could start out with something simple
>and high level and
>then expand.
>
>Some basic "stream of consciousness"scenarios from
>individuals would help me get
>started.
>
>Best regards,
>Denise
>
>
> (06)
>
> Nicolas F
>
> Rouquette
>
> <nicolas.rouque
> To
> tte@xxxxxxxxxxx "[ontolog-forum]"
>
> v>
> <ontolog-forum@xxxxxxxxxxxxxxxx>
>
> Sent by:
> cc
> ontolog-forum-b
>
> ounces@ontolog.
> Subject
> cim3.net Re: Tagging Ontolog
>content [was - Re:
> [ontolog-forum]
>ONTOLOG forum podcast]
> (07)
>
> 02/07/2006
>
> 02:49 PM
>
> (08)
>
> (09)
>
> Please respond
>
> to
>
> nicolas.rouquet
>
> te@xxxxxxxxxxxx
>
> ; Please
>
> respond to
>
> "[ontolog-forum
>
> ] "
>
> <ontolog-forum@
>
> ontolog.cim3.ne
>
> t>
>
> (010)
>
> (011)
>
>
>
>
>
>Denise,
>
>Taking the architecture "high-road" as you suggest would
>be indeed a
>wise strategy.
>However, this would require someone investing enough time
>& energy to do it.
>Are you volunteering to do that?
>
>A shortcut may be to demonstrate what is technologically
>easy to demonstrate
>in this context as a way to collaboratively jumpstart the
>important
>"high-road"
>discussion on architecture. A simple technology
>demonstration may
>illustrate
>only a small part of this architecture and it might, if
>misused, mislead
>the perspective
>on what the "high-road" architectural principles are. This
>is a risk;
>however,
>I believe it also provides a point of reference for
>enlightened
>individuals like
>you to explain the merits & deficiencies of that example
>w.r.t. the
>"high road"
>architecture principles you believe in more easily than if
>you have to
>do the
>architecture thinking.
>
>A technology demo/example seems to me a more expedient way
>to explain
>what are the key architecture principles important to the
>semantic
>integration
>of diverse kinds of information. Without examples, I'm
>afraid we could
>have a very abstract, dry and boring discussion with lots
>of hand-waving,
>conjecturing, posturing, etc....; i.e., a good recipe for
>a boring
>discussion.
>
>-- Nicolas.
>
>dbedford@xxxxxxxxxxxxx wrote:
>
>> All,
>>
>> You need to do the second session before you do the
>first. This
>> challenge is a 'thinking' problem more than it is a
>technology
>> problem. And, it is also an architecture and a workflow
>issue.
>> Having worked on these problems for about 15 years now I
>can tell you
>> that there is no technical solution that doesn't require
>a supporting
>> architecture, careful planning, and an investment of
>time. To
>> integrate structured/semistructured/unstructured
>information, you need
>> a different kind of architecture.
>>
>> You can succeed big time, though, and gain real
>productivity and
>> access improvements if you take a thoughtful approach.
>>
>> Best regards,
>> Denise
>>
>>
>>
>> Best regards,
>> Denise
>>
>>
>> -----ontolog-forum-bounces@xxxxxxxxxxxxxxxx wrote: -----
>>
>> To: "[ontolog-forum]"
><ontolog-forum@xxxxxxxxxxxxxxxx>
>> From: "Peter P. Yim" <peter.yim@xxxxxxxx>
>> Sent by: ontolog-forum-bounces@xxxxxxxxxxxxxxxx
>> Date: 02/03/2006 03:52PM
>> Subject: Tagging Ontolog content [was - Re:
>[ontolog-forum]
>> ONTOLOG forum podcast]
>>
>> > [Max] the new podcast schema (I think 2.0) which
>allows you
>> > to add some forms of annotations and pointers
>during the
>> > podcast. So for instance you can add different
>images so
>> > that as the podcast plays, the image (we have none
>for now)
>> > that is displayed in iTunes would change.
>>
>> [ppy] re:image - the speakers' slide(s) would be a
>natural candidate!
>>
>> > [Max] We should brainstorm and discuss some more.
>> &g ; Nicolas/Peter (others) any suggestions on how
>to proceed?
>> > I am happy to continue via email and then do some
>live
>> > discussion.
>>
>> [ppy] ... something more concrete: how about
>organizing a
>> two-part technical discussion series,
>>
>> (a) the first session surveying the relevant
>technology that's
>> "out there" now: ... topic maps, swoop, folksonomy,
>UIMA ... etc.
>> (especially those for unstructured and
>semistructured content), and
>>
>> (b) the second session devoted to a discussion on
>what we want to
>> do (with the Ontolog content), and how we could
>proceed ... with
>> this possibly leading up to another Ontolog project.
>>
>> We can do (a) as early as Feb. 16 or Feb. 23, 2006,
>if both of
>> you (Nicolas and Max) are willing to co-moderate
>(and, obviously,
>> co-organize) it. The challenge would be: Can you
>pull together
>> the real gurus for that in short order? If the
>answer it yes,
>> that would be just great!
>>
>> Additional volunteers and input from everyone are
>welcomed.
>>
>> Cheers. =ppy
>> --
>>
>>
>> Michael Maximilien wrote Fri, 3 Feb 2006 11:52:33
>-0800:
>> >
>> > > Max: is podcasting ontolog forum discussions
>sufficiently
>> practical &
>> > > reasonable
>> > > in terms of resource demands to do on a
>systematic basis?
>> >
>> > Yes.
>> >
>> > > If yes, then this raises a number of
>interesting
>> possibilities that
>> > > could help us better "digest our own ontology
>food"
>> >
>> > You mean 'dogfooding', I like that :-)
>> >
>> > > For example, Nicola Guarino used the famous
>"red rose"
>> example in his
>> > > presentation of DOLCE
>> > > (see
>>
>
http://ontolog.cim3.net/cgi-bin/wiki.pl?ConferenceCall_2006_02_02)
>> >
>> > I just added his talk to the podcast.
>> >
>> > (Just downloaded it so that I listen to it over
>the week- nd :-)
>> I had
>> > to miss most of it...)
>> >
>> > > Wouldn't it be great then to be able to
>annotate the podcast w/ a
>> > > pointer to an example ontology built (for
>example)
>> > > using Swoop from the public URL of DOLCE in a
>way that
>> includes "back"
>> > > annotation links to the podcast
>> > > to establish a 3-way link among
>> >
>> > Yes. This is very interesting indeed.
>> >
>> > In that same spirit but more from a engineering
>point of view, I
>> would
>> > also point the group (I am sure you are aware of
>this) to the UIMA
>> > (http://uima-framework.sourceforge.net/) work and
>also the new
>> podcast
>> > schema (I think 2.0) which allows you to add some
>forms of
>> annotations
>> > and pointers during the podcast. So for instance
>you can add
>> different
>> > images so that as the podcast plays, the image (we
>have none for
>> now)
>> &g ; that is displayed in iTunes would change.
>> >
>> > What you are suggesting is a more generic version
>of that...
>> very cool.
>> >
>> > > Annotations need to be "first-class" things
>that can
>> themselves be
>> > > annotated.
>> > > In this case, the above annotation is an
>example of the
>> distinction
>> > > between descriptions (3) and situations (1) in
>DOLCE.
>> >
>> > Yes.
>> >
>> > BTW, (you probably know this as well) but here is
>a pointer to a
>> paper
>> > by Tom Gruber on folksonomy and ontology where he
>begins a model
>> for (012)
>> > representing tagging and essentially making them
>"first-class".
>> >
>>
>
>http://tomgruber.org/writing/ontology-of-folksonomy.htm
>> >
>> > (I believe, in addition to Peter, two or three
>others may have
>> suggested
>> > Tom's paper to me---including John Domingue when I
>met him in A'dam
>> > during ICSOC last November)
>> >
>> > > The mindswappers have shown that doing
>annotations and
>> referencing to
>> > > subsets of ontologies is technically doable.
>> > > One relevant question is whether the
>> tools/mechanisms/infrastructure to
>> > > do it could be deployed as an on-going service
>for this forum.
>> >
>> > Interesting. It's doable. There are a variety of
>possibilities,
>> > including using/creating services that allow
>content to be
>> > tagged/annotated and then composing/mashing that
>with the Wiki. It
>> > might take some engineering effort but definitely
>doable...
>> >
>> > We should brainstorm and discuss some more.
> Nicolas/Peter
>> (others) any
>> > suggestions on how to proceed? I am happy to
>continue via email
>> and
>> > then do some live discussion.
>> >
>> > Best,
>> >
>> > E. M. (Max) Maximilien, Ph.D.
>> > IBM Almaden Research Center
>> > San Jose, C USA
>> > maxim@xxxxxxxxxx
>> > Homepage: http://maximilien.org
>>
>>
>> > ontolog-forum-bounces@xxxxxxxxxxxxxxxx wrote on
>02/03/2006
>> 11:15:29 AM:
>> >
>> > > Thanks Peter for the heads-up/reminder.
>> > >
>> > > Max: is podcasting ontolog forum discussions
>sufficiently
>> practical &
>> > > reasonable
>> > > in terms of resource demands to do on a
>systematic basis?
>> > >
>> > > If yes, then this raises a number of
>interesting
>> possibilities that
>> > > could help us better "digest our own ontology
>food"
>> > > with related technology support (e.g., Swoop's
>annotation
>> support &
>> > > Annotea server, publicly available ontolgies
>> > > w/ sufficiently axiomatized content adequate
>for reasoning
>> analysis --
>> > > e.g., SUMO, DOLCE, ...)
>> > >
>> > > For example, Nicola Guarino used the f mous
>"red rose"
>> example in his
>> > > presentation of DOLCE
>> > > (see
>>
>
http://ontolog.cim3.net/cgi-bin/wiki.pl?ConferenceCall_2006_02_02)
>> > >
>> > > As Nicola emphasized at the end of his talk,
>small examples
>> are really
>> > > important to convey
>> > > and discuss important ideas like that of being
>able to
>> precisely capture
>> > > the subtle differences
>> > > on the various meanings of "red" in "red rose".
>> > >
>> > > Wouldn't it be great then to be able to
>annotate the podcast w/ a
>> > > pointer to an example ontology built (for
>example)
>> > > using Swoop from the public URL of DOLCE in a
>way that
>> includes "back"
>> > > annotation links to the podcast
>> > > to establish a 3-way link among
>> > >
>> > > (1) a subset of the exampl ontology that
>illustrates one of the
>> > > meanings of "red"
>> > > (2) a segment of the presentation where Nicola
>explains that
>> particular
>> > > meaning
>> > > (3) the description of that meaning in DOLCE
>> > >
>> > > Annotations need to be "first-class" things
>that can
>> themselves be
>> > > annotated.
>> > > In this case, the above annotation is an
>example of the
>> distinction
>> > > between descriptions (3) and situations (1) in
>DOLCE.
>> > >
>> > > The mindswappers have shown that doing
>annotations and
>> referencing to
>> > > subsets of ontologies is technically doable.
>> > > One relevant question is whether the
>> tools/mechanisms/infrastructure to
>> > > do it could be deployed as an on-going service
>for this forum.
>> > >
>> > > -- Nicolas.
>>
>>
>> > > Peter P. Yim wrote Wed, 01 Feb 2006 06:44:37
>-08 0:
>> > >
>> > > > Thank you very much, Max. This is great!
>> > > >
>> > > > See announcement at the Ontolog:WikiHomePage
>-
>> > >
>>http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nidJMZ
>> > > > (and the ensuing details under the "Events"
>section.)
>> > > >
>> > > > Also, please refer to our discussion at last
>week's call
>> relating to
>> > > > this matter -
>> > > >
>>
>
>http://ontolog.cim3.net/cgi-bin/wiki.pl?ConferenceCall_2006_01_26#nidJKM
>> > > >
>> > > > In particular, there is continuing work to
>maintain/enhance the
>> > > > podcast, and we need some volunteers (contact
>Max for details).
>> > > >
>> > > > More inter stingly, an initiative that
>Nicolas Rouquette
>> alluded to
>> > > > before is now re-surfacing as a potential
>project for "tagging
>> > > > multimedia and other Ontolog content for
>podcast & semantic
>> search."
>> > > > ... let's discuss this. ... Better still,
>maybe we can even
>> devote a
>> > > > technical discussion session for it (Nicolas
>and Max, you
>> want to
>> > > > co-organize this?)
>> > > >
>> > > > Thanks & regards. =ppy
>> > > > --
>>
>>
>> > > > Michael Maximilien wrote Wed, 25 Jan 2006
>12:58:23 -0800:
>> > > >
>> > > >>
>> > > >> Hi,
>> > > >>
>> > > >> 1) Podcast available on ONTOLOG forum Wiki.
>> > >
>>
>
>>http://ontolog.cim3.net/file/resource/podcast/ontolog-podcast.xml<
>> r>> > >>
>> > > >> 2) Modified Wiki homepage to announce ---
>feel free to
>> edit, add
>> > > >> images, or change text
>> > > >>
>> > > >> 3) Added all 2005 episodes -- I will
>gradually add more as
>> time
>> > permits
>> > > >>
>> > > >> 4) Registered podcast with iTunes and Yahoo
>podcast --
>> soon you will
>> > > >> be able to search and subscribe directly
>from these sites
>> > > >>
>> > > >> 5) Tried RSS feed on iTunes and Safari ---
>works fine
>> > > >>
>> > > >> 6) Removed RSS feed from my site
>> > > >>
>> > > >> Thanks to Peter Yim for various assistance
>and making this
>> happen.
>> > > >> Send me and Peter any feedback or better yet
>respond to
>> this email.
>> > > >>
>> > > >> If you have problem us ng then let me know.
> Take care,
>> > > >>
>> > > >> E. M. (Max) Maximilien, Ph.D.
>> > > >> IBM Almaden Research Center
>> > > >> San Jose, CA USA
>> > > >> maxim@xxxxxxxxxx
>> > > >> Homepage: http://maximilien.org
>>
>> > > >> ...[snip]...
>>
>
_________________________________________________________________
>> Message Archives:
>http://ontolog.cim3.net/forum/ontolog-forum/
>> Subscribe/Unsubscribe/Config:
>>
>
http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
>> Shared Files: http://ontolog.cim3.net/file/
>> Community Wiki: http://ontolog.cim3.net/wiki/
>> To Post: mailto:ontol g-forum@xxxxxxxxxxxxxxxx
>> <mailto:ontolog-forum@xxxxxxxxxxxxxxxx>
>>
>>
>
>_________________________________________________________________
>Message Archives:
>http://ontolog.cim3.net/forum/ontolog-forum/
>Subscribe/Unsubscribe/Config:
>http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
>Shared Files: http://ontolog.cim3.net/file/
>Community Wiki: http://ontolog.cim3.net/wiki/
>To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx
>
>
>
>
>_________________________________________________________________
>Message Archives:
>http://ontolog.cim3.net/forum/ontolog-forum/
>Subscribe/Unsubscribe/Config:
>http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
>Shared Files: http://ontolog.cim3.net/file/
>Community Wiki: http://ontolog.cim3.net/wiki/
>To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx
> (013)
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Subscribe/Unsubscribe/Config:
http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx (014)
|