Great questions, Owen, and great response, Peter! (01)
I keyed into the term "stovepipe". The comments I will offer here got
their start while working with Peter and Adam Cheyer at VerticalNet.
There, we built an upper ontology under the guidance of Leo Obrst for
the express purpose of hanging "vertical" communities (stovepipes by
another name) under that "triangle" (the diagrammatic illustration of an
upper always looks like a triangle). It started to occur to me that, as
communities began to evolve, creating "concepts" within their, um,
stovepipes, perhaps some of those concepts could be made usable (read:
made visible) to other communities. Thus was born the notion of
In light of world events, I have evolved that idea to one I call
"cultural federation" and it is cultural norms that, I think, drive
Owen's inquiry here. In my view of "cultural federation", no cultural
norm should be extinguished in favor of some foreign norm; if anything,
all cultural norms should be encouraged (massive caveats lurking behind
that one), that is, aided in the Engelbartian NIC sense. But, to the
extent that many of the cultural "concepts" which make up cultural norms
within COPs overlap, and I would say, overlap on a "same subject" basis,
then the artifacts which represent those subjects become shared
conceptual resources, no matter how those artifacts are used. (03)
Borrowing from a paper I just submitted to the Semantic Desktop 06
workshop, I would offer the suggestion that promiscuous semantic
federation of cultural knowledge assets is to be encouraged.
Justification for that assertion comes from various anthropological
studies which suggest that sexual promiscuity was good for the gene
pool; it follows (my opinion, your mileage might vary) that memetic
promiscuity might be good for the meme pool. (04)
When you merge same-subject conceptual representations, nobody forces
anybody to even so much as look at the totality of the merged product;
folks are free to roam a representation space (e.g. library) carrying
any ontological biases they might have. As Patrick Durusau and I argued
in our Ontolog talk , even casually looking at representations of
other cultures within a same-subject view of all representations opens
the door to chance discovery; sometimes you see things you just didn't
know existed. http://del.icio.us/ is proving that on a daily basis.
That's a statement that's downright difficult to grok until you actually
get a free account there and play with it. For more on that, see Jon
Udell's blog  (05)
All of that to suggest that a public information commons that serves
CoPs of all kinds can, at once, leave each CoP to its own devices, and
federate them in ways that can be shown to be as beneficial as an uberNIC. (06)
I'm tempted to close with "Promiscuous Semantic Federation" rocks!, but
I won't ;) (07)
 http://weblog.infoworld.com/udell/gems/delicious.html (08)
Peter P. Yim wrote:
> Let me take a crack at your questions (purely as an attempt to get more
> dialog going) ...
> ... I'll take it from a fictitious position of someone trying to sell a
> hardcore-XMLmer on Ontologies or Ontolog (much the same way that XMLmer
> had to sell the hardcore-Database guys on why they would need XML). I'll
> attempt only one section at a time ... and hope that others will pick up
> and continue to answer the rest of your questions in the interim.
> >> [OA] 1) How can we avoid forcing folks like me to subscribe
> >> to yet another group, listserv, "portal" or otherwise named
> >> "stovepipe" system in order to share information with those
> >> with whom they hold interests in common, however fleeting or
> >> narrowly focused those interests might be?
> >> For example, can an XML schema be specified enabling folks
> >> to describe their own interests on *any* site (e.g., their
> >> own, personal Web sites) anywhere on the Web, thus avoiding
> >> the need for them to subscribe to anyone else's site, while
> >> intermediary sites could index, automatically establish
> >> linkages, and enable productive syntheses of information
> >> along specialized lines of interest?
> [ppy] Let's look at this from two angles. The social aspect of
> participation in a community of practice (CoP), and the mechanism of
> "sharing information" in the scenario you described.
> First off, the CoP (see more about what we mean by that at:
> http://ontolog.cim3.net/cgi-bin/wiki.pl?CommunityofPractice, which
> Ontolog claims to be a virtual version of one) is a "small group of
> people who've worked together over a period of time. Not a team, not a
> task force, not necessarily an authorized or identified group. They are
> peers in the execution of 'real work.' What holds them together is a
> common sense of purpose and a real need to know what each other knows."
> ... I, for one, find this setting to be rather attractive in handling
> "wicked problems" (ref: http://www.cognexus.org/id42.htm
> http://en.wikipedia.org/wiki/Wicked_problems or
> http://www.poppendieck.com/wicked.htm), and I purport that most of us
> are actually working in the "wicked problems" (as opposed to "tame
> problems") space. Most of the time, we can hardly articulate the problem
> ... let alone going straight off to hammer out the solution. We need
> this kind of organic work environment and ecology to "develop" what may
> become viable solutions to those problems.
> Participation in a CoP should be more akin to hanging out at the tavern,
> rather than going to class. "Forcing" would be too strong a word for the
> As for the mechanism ... I believe what you are asking for is actually
> getting there, as web 2.0 implementations become more pervasive. At
> Ontolog, we do things like RSS and podcast too!
> One caveat, though, when you talk about "sharing information" and start
> making references to notions like "interests," "productive syntheses of
> information," ... etc. you are talking "sharing knowledge" and
> "knowledge reuse," rather than just going after mechanisms that helps
> ship bits (or messages) around. One more enabler seems to be in order,
> especially if one wants to take the human out of the loop (and let the
> machine-to-machine process take on the workload) as much as possible ...
> and that would be our "ontology".
> That additional layer of abstraction, would potentially help us discern
> (using your example), when I say my interest is in "CIM," whether I am
> actually interested in "Computer Integrated Manufacturing," "Common
> Information Model," "Customer Information Management," or "Curtis
> Institute of Music," or the hundreds of other possibilities that "CIM"
> might denote. ... And, if the system already 'knows' its the first of
> those, then it could also be 'aware' that I might, at the same time, be
> interested in "automation," "AS/RS," "ERP," "process simulation," ...
> and a whole bunch of other things (noting that none of which even share
> the same search word/term with what the CIM acronym stood for.)
> Therefore, the short answer to your question:
> (a) I think *working this out in a CoP setting is a good idea* (we're
> not forcing anyone), and
> (b) to enable what you are looking for, we need to be able to properly
> represent 'semantics' in our systems, and, to that end, *the need for
> ontologies* seems to be in order.
> Cheers. =ppy
> Peter P. Yim wrote Mon, 14 Aug 2006 12:23:43 -0700:
>> Thank you, Owen.
>> Folks ... Owen Amber (DoI, XML-CoP) came back with this set of thought
>> provoking, and actually, rather challenging questions.
>> Each and everyone of Owen's questions calls for some deep thinking,
>> some form of answer/solution, but (evidently, more importantly to
>> Owen) some action to follow.
>> ... comments, suggestions, insights, solicited. Extending the real
>> time discussion to this asynchronous platform, let's see if our
>> community can 'collectively' take a crack at Owen's challenge!
>> Cheers. =ppy
>> Owen_Ambur@xxxxxxxxxxx wrote Mon, 14 Aug 2006 08:49:52 -0400:
>>> Peter, I'm not sure I can afford to be subscribed to more groups or
>>> stovepipe IT systems, and I am skeptical about the potential
>>> of my involvement in the ontolog forum. However, since you and Duane
>>> think it might be worthwhile, I am willing to give it a try.
>>> Aside from the issue of what actions folks like me may wish to consider
>>> taking as a result of dialog on the telecon last week, here are some
>>> other issues I'd like to explore with anyone who may share my
>>> interest in doing so:
>>> 1) How can we avoid forcing folks like me to subscribe to yet another
>>> group, listserv, "portal" or otherwise named "stovepipe" system in
>>> order to share information with those with whom they hold interests
>>> in common, however fleeting or narrowly focused those interests might
>>> For example, can an XML schema be specified enabling folks to describe
>>> their own interests on *any* site (e.g., their own, personal Web sites)
>>> anywhere on the Web, thus avoiding the need for them to subscribe to
>>> anyone else's site, while intermediary sites could index,
>>> automatically establish linkages, and enable productive syntheses of
>>> information along specialized lines of interest?
>>> 2) How can the term "ontology" be operationally defined in terms of
>>> what folks like me might be able to do -- on our own Web sites, with
>>> no training nor any specialized tools -- to contribute to realization
>>> of the vision of the semantic Web?
>>> BTW, while I am far from being an ontological expert, from what I
>>> have seen and heard, my working definition of the term "ontology" is
>>> "someone else's top-level taxonomy that they are trying to impose on
>>> 3) On the telecon last week someone (I think it was the guy from
>>> Zapthink) said one of his customers decided not to use XML for one of
>>> their applications because there was no need to deal with the
>>> excessive overhead
>>> associated with the verbosity of the XML coding. While it is easy to
>>> understand why compression and other means might be used to
>>> accelerate the transmission of the electrons, the separation of
>>> content from its
>>> presentation has much to do with the fact that so many of our business
>>> processes are still so bound up in paper. Intuitively as well as
>>> speaking, people know they cannot trust systems whose records lack
>>> integrity, including presentational integrity. The speed with which
>>> electrons can be transmitted is often not the most important business
>>> issue, indeed far from it.
>>> Thus, I'd be curious to know whether the data being transmitted in the
>>> referenced application is important and, if so, whether it ever needs
>>> to be shared with anyone else, e.g., partners, auditors,
>>> shareholders, or other stakeholders. If not, how can it be
>>> important? If so, do the benefits of
>>> accelerated internal transmission offset the costs associated with the
>>> inability to share the data readily with others? Presumably so, but I'm
>>> curious to know how the records management (and sharing) requirements
>>> being addressed.
>>> 4) Finally and perhaps most importantly, are there reusable components
>>> and/or specifications related to ontologies for which the CIO Council's
>>> ET.gov site/process might productively be used to build .gov
>>> communities of
>>> practice to foster implementation and incorporation in the Federal
>>> Enterprise Architecture (FEA) Technical Reference Model (TRM) or Service
>>> Component Reference Model (SRM) and/or the Federal Transition Framework
>>> (FTF) Catalog, for potential Governmentwide usage?
>>> Note: Stage 1 of the ET.gov site/process implements the principle
>>> outlined in item 1 above. http://et.gov/stage1.htm Folks can post
>>> their valid XML
>>> instance documents anywhere on the public Web. They are not required to
>>> "subscribe" to or otherwise re-create their digital personas at ET.gov.
>>> Indeed, to avoid implications associated with the Privacy Act, we do not
>>> store any personally identifiable information on the site. If
>>> wish to provide such information, it is stored on their own site or the
>>> site or the site of any intermediary service they may choose to use.
>>> If you see any prospects for mutually productive pursuit along any of
>>> these lines, I will be pleased to explore opportunities for
>>> collaboration with anyone who may be interested in actions (small,
>>> manageable steps) that we might take together as well as individually.
>>> "Peter P.
>>> om> To
>>> Duane Nickull
>>> <dnickull@xxxxxxxxx> 08/11/2006
>>> 04:59 cc
>>> PM Owen Ambur/PIR/OS/DOI@DOI, Rex
>>> Subject Re: [ontolog-forum]
>>> Proceedings of the Ontolog
>>> Discussion - Ontologies & SOA
>>> - Moderators: DuaneNickull &
>>> Rex Broooks - Thu 2006.08.10
>>> Great idea, Duane.
>>> Owen, let me subscribe you to the [ontolog-forum] list so that
>>> you can post to it. =ppy
>>> Duane Nickull wrote Fri, 11 Aug 2006 13:14:12 -0700:
>>>> I would like to actually suggest that instead of us answering that
>>>> perhaps Owen can post this request to the Ontolog Forum list. I
>>>> think others have a similar interest.
>>>> On 8/11/06 1:02 PM, "Peter P. Yim" <peter.yim@xxxxxxxx> wrote:
>>>>> Hi Owen,
>>>>> It's just great that you were able to join us (at least for the
>>>>> good part of the session) yesterday.
>>>>> I'll let Duane or Rex answer you, specifically (although
>>>>> identifying reusable components probably wasn't one of the
>>>>> discussion focus yesterday ... it could well be for another session!)
>>>>> I look forward to more collaboration with you and your
>>>>> communities in the future.
>>>>> Best regards. =ppy
>>>>> Owen_Ambur@xxxxxxxxxxx wrote Fri, 11 Aug 2006 15:07:29 -0400:
>>>>>> Peter, Duane, Rex & Brand, I had to drop off the telecon yesterday
>>>>>> at 2:30. Were any conclusions drawn that might be subject to
>>>>>> action by folks like me?
>>>>>> For example, were any reusable components or relevant technical
>>>>>> specifications identified that might be candidates for the ET.gov
>>>>>> site/process? http://et.gov/
>>>>>> BTW, Peter, I like the way you have introduced Rex in your subject
>>>>>> line. Kinda like Ed McMahon used say when he introduced Johnny
>>>>>> Carson -- here'sRex Brooooks!
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx (010)