Doug Engelbart tells the story about how he once got dinged for not
defining knowledge. I choose not to go there in the flash I am having
over this post. Chris asks good questions, and often gives good answers,
particularly when he surfaces to comment at all. In my mind, the whole
of all of Chris's questions conflate to one, just one thing, and that
thing is information. Information gets munged into knowledge which
eventually turns into wisdom, and all of that only in wetware, not in
hardware or software. Enough of the philosophical stuff. (01)
A knowledge base had otta better be able to capture, some would say
"represent" information in a way that the syntaxes and semantics are
captured such that wetware can acquire and interpret it. Nothing more,
nothing less (my opinion, and your mileage may vary, skilled driver on a
closed course and all that). Are there characteristics I would like to
see in a knowledge base? Yup. Scalability, evolvability, not suspect to
schema creep, capable of capturing information in a wide variety of ways
(e.g. symbols, probabilities, fuzzies, whatever). (02)
Knowledge bases go together hand in hand, imho, with content bases, that
is, content management systems. That's because the universe of discourse
captured in a knowledge base is likely busy pointing to occurrences
(topic map speak) out in content land. (03)
Knowledge management falls out of that, when the architectures of both
the kb and the cb facilitate the process, both automated and human
directed, which connect the dots, provoke the aha! moments, and so forth. (04)
In some sense, I agree with Pollard that specialty tools survive best in
constrained universes. But,I don't think that we have exhausted the
possibilities. I am profoundly persuaded to like the Knownspace project
for its efforts to find ways to, um, represent information in a scalable
and extensible way, not at all subject to schema creep and all that, and
for its underlying religious entities: simpletons. Tiny agents that do
simple things. That also plays to the scalable and evolvable architecture. (05)
MIT, under financial resources of HP, have come up with DSpace
www.dspace.org, and it is a content management system, complete with
workflow engines, tons of metadata, and all that. Seems to me that,
given a few more functionalities, it would make a damn fine knowledge
base and, hence, knowledge management system as well. (06)
In topic map land, and in diagrams I once drew during the unrev hayday,
there were, essentially, three bubbles. One for content, one for
knowledge, and one for context -- the human in the loop. They all go
together. They cannot be separated. That's what happens, I think, when
entities produce software for content management, or knowledge
management, but not together. (07)
Enough said. Plenty of troll bait, what?
Jack (08)
cdent@blueoxen.org wrote: (09)
>Crap I sent this without finishing it...
>
>On Mon, 15 Dec 2003, Chris Dent wrote:
>
>[snip]
>
> > My own feeling is that a knowledge base is a necessary condition
> > for a knowledge management system but is not sufficient.
> >
> > A knowledge base is, to my mind, a collection of explicit
> > knowledge made accessible. And perhaps the tools to manage
>
>...the content.
>
>A knowledge management system, on another hand, is a suite of
>tools and processes that encourages the transfer of tacit
>knowledge into explicit and accessible form. It's quite likely it
>would include a knowledge base but also it would have such things
>as diaries or blogs that can be used for later markup (for reuse)
>as well as people finders.
>
>Dave Pollard has some descriptive info
>
> http://blogs.salon.com/0002007/2003/03/18.html#a123
>
>that is food for thought.
>
>Complicating the matter in the IU context is that people there
>seem to be conflating Knowledge Management System with Content
>Management System. So we get the following:
>
>* There exists a knowledge base of questions and answers
>* Some people think it is a knowledge management system
>* Some people think a knowledge managment system is a content
> management system.
>* Being university folk, they want to centralized control
>* The decide, hey we've got a kms in the kb, let's centralize
> content management there and deliver it from there to all IU
> web sites.
>
>That's a problem when the existing kb is purpose built for data
>that is structured as questions and answers. It doesn't think in
>terms of pages, documents, or collections.
>
>So, to review the questions (which I've may not have stated
>explicitly yet):
>
>* What are the characteristics of a knowledge base?
>
>* What are the characteristics of a knowledge management system?
>
>* What are the characteristics of a content management system?
>
>Thanks for any thoughts and opinions.
>
>
> (010)
--
This message is archived at: (011)
http://collab.blueoxen.net/forums/cgi-bin/mesg.cgi?a=tools-yak&i=3FDE90A8.7000406@thinkalong.com (012)
|