Re: Intro -- Ian Hosking LO1866

Michael McMaster (Michael@kbddean.demon.co.uk)
Thu, 29 Jun 1995 17:15:34 +0000

Replying to LO1822 --

The ideas of a "knowledge data base" come up frequently in
corporations interested in learning. Most I think are seriously
flawed. There may be gold in here somewhere but, until I've seen
clear distinctions and theory around this idea, I'd worry that a lot
of money can be poured down this particular drain.

I recall that GE was one of the largest investors in data base
building back in the early 80s (or so) and spent enormous amounts for
which they got little value - except possibly some learning about
such ventures. But the price was very high as I understand it.

This was before the phrase "knowledge data base" was around.

I assume that the distinction _knowledge_ data base refers to a data
base of knowledge or knowledge sources rather than to a data base of,
say, information. The problem here is that "knowledge" is not
operationally distinguished. Knowledge, in my operational
distinctions (and grounded in biology - a la Varela - and philosophy
- a la interpretive/linguistic), refers to the integrations of
information into patterns that are available for action. These are
not discrete bits but whole patterns that are interconnected with
other patterns. In particular, knowledge is directly related to
action or action potential. We can say we "know something" when the
appropriate actions occur in the circumstances that call for that
action.

The point of the depth here is make explicit that knowledge is
context dependent. That is, its very existence is context dependent.
I can't know that I know until the context has it occur.

This reveals a particular problem for a "knowledge data base".
Knowledge doesn't exist that way. BUT - maybe useful work can be
done nonetheless. The knowledge of an organisation isn't a data base
and doesn't exist in a data base any more than the knowledge of an
individual does. I maintain that there is "organisational knowledge"
and that this is not the equivalent of the sum (or any other
mathematical term) of the individual knowledge.

If the "knowledge data base" is attempting to get at this in a
pragmatic way, then there are some things that we can do. I think
we'll have to use principles of complex adaptive systems to be
effective and these are far from well developed. However, there are
some directional possibilities. One of these is to build a system
that creates pathways to possible patterns of knowledge based on the
request/demand for that knowledge rather than its "already"
existence.

In one case, a company was trying to record what each person knew
_and was willing to share_ and they had accumulated quite an
interesting display. However, it turned out that the knowledge
wasn't forthcoming in many instances where it was wanted. It also
turned out that many people were willing to share knowledge in some
contexts and not in others and the system had no way of sorting these
out. So far, that particular system is barely functional and taps
little of the knowledge that everybody knows is there.

We need to rurn our attention to design of connections and what calls
the knowledge forth without having any idea of what that knowledge is
or what it will be when called forth.

--
Michael McMaster
Michael@kbddean.demon.co.uk