RE: [xtm-wg] Knowledge management claims re XTM (and Topic Maps.. .)

["Dale Hunscher" <dale@supportability.com>]

Situation semantics, eh? Sounds intriguing. I will take a look at that, too.

Your CBR scenario rang a bell... it's been a while since I followed the AI
community (now in diaspora of course) so I'm definitely out of touch with
their latest and greatest.

The example you gave shows what I think is the most practical immediate use
for Topic Maps and applications based on it. TM could be very useful, along
with *limited* semantic analysis, as an intelligent assistant to researchers
in the companies for whom I design editorial and publishing systems in the
pharmaceutical arena. The goal would be intelligently to filter and
prioritize search results. Automatic linking of occurrences to TMs would be
of great importance, but one would always want to be able to examine the
reasoning behind decisions made by the software.

-----Original Message-----
From: Wrightson, Ann [mailto:Ann.Wrightson@sweetandmaxwell.co.uk]
Sent: Thursday, August 24, 2000 4:11 AM
To: 'xtm-wg@egroups.com'
Subject: RE: [xtm-wg] Knowledge management claims re XTM (and Topic
Maps.. .)


Yes, and there's more. (I tend to prefer situation semantics, but that's a
(very) minority view!)

Steve Pepper had a suitable KE-principles-review book with him at Paris, but
I forget title/author. A glance at the Cambridge UP "Knowledge Engineering"
journal will show you how heavyweight "publishable" stuff in KE proper tends
to be - but there are also a number of areas where KE is applied, which is
less "heavy" - and that's the sort of stuff we'll need, for which picking up
what's in the-book-I-forget or a similar graduate-level textbook will be
AOK.

A quick back-of-an-envelope example where I think we can get v. natural
synergy is case-based reasoning, where TM could link formal case
characterizations (which identify topics) with examples (occurrences - eg
documents which involve examples of such cases, within one or more data
mines...), and then also link to descriptions of and interrelationships
between particular ways of reasoning with the formal case descriptions
(there are lots) - and the whole would then be used by some case-based tool
using the TM as its data finder, at some point presenting the links to
data-mine resources to a human as appropriate (eg as part of the display of
a case identified by some computation).

Cheers

Ann W.


> -----Original Message-----
> From: Dale Hunscher [mailto:dale@supportability.com]
> Sent: 24 August 2000 14:49
> To: xtm-wg@egroups.com
> Subject: RE: [xtm-wg] Knowledge management claims re XTM (and Topic
> Maps.. .)
>
>
> Hello Ann,
>
> Is it possible to provide some pointers to Web links that
> would help provide
> some concise definitions for these terms as understood by the
> KM folks these
> days? Or failing that, any useful reading in the form of a
> book or two?
> Question from a conference-speaker-to-be this Fall who wants
> all the toes he
> steps on to be deliberate targets... ;-)
>
> Am I correct in thinking that the biggest real deficiencies
> of topic maps
> for KM purposes are the lack of real semantics, i.e., hooks that would
> support analysis of relationships in terms of symbolic
> logic--equivalence,
> transitivity, etc. Or is there more? Of course topic maps
> themselves are not
> tools for inferencing per se, but the goal would be to make
> sure they carry
> enough information to *support* inferencing, as you
> indicated. (Detailed
> reply not necessary... something along the lines of "No ,
> there's more see X
> for more details" would be more than sufficiently illuminating...)
>
> Thanks,
>
> Dale
>
> -----Original Message-----
> From: Wrightson, Ann [mailto:Ann.Wrightson@sweetandmaxwell.co.uk]
> Sent: Thursday, August 24, 2000 2:26 AM
> To: 'xtm-wg@egroups.com'
> Subject: RE: [xtm-wg] Knowledge management claims re XTM (and Topic
> Maps.. .)
>
>
> Eric....
>
> The main reason I was unspecific was that from the
> presentations I heard, I
> didn't think that *in the context of an XML conference*,
> anyone had actually
> spoken out of turn; also because of the impression I had that
> the annoyance
> of the folks concerned had built up over the session rather than being
> caused by any one speaker. "The problem" if you like was the
> sensitivities
> of the KE folks! - in particular, "the problem" was their
> interpretation as
> technical terms, and so as specific technical claims of what
> Topic Maps
> could "provide", some words and phrases which I guess were
> intended as more
> general descriptions (& eye-openers for XML-folks) of the
> potential uses of
> Topic Maps. (For example, I remember one forceful statement
> that a bunch of
> associations and instances on a diagram plus some informal
> typing didn't
> amount to a "semantic network" - and that such a thing would
> never pass
> review in any of "their" conferences - which last may even be
> the central
> underlying emotional point.)
>
> The particular phrases/words which I believe from that
> experience that we
> need to use with due care are:
> - semantic net or network
> - knowledge model
> - inference (especially the difference between being capable
> of supporting
> inference, and being a sufficient agent or means of inference
> in itself)
>
> I should also say that I believe many of us are already being
> careful and
> exact... and that this is not the first time I've seen the
> re-purposing of
> words & phrases for precise uses in computer science leading to
> misunderstandings.
>
> Does this answer?
>
> Cheers
>
> Ann W.
>
> > -----Original Message-----
> > From: Eric Freese [mailto:eric@isogen.com]
> > Sent: 23 August 2000 15:31
> > To: XTM Working Group
> > Subject: RE: [xtm-wg] Knowledge management claims re XTM (and Topic
> > Maps...)
> >
> >
> > Ann:
> >
> > Perhaps you could illuminate us with what claims caused these
> > people so much
> > distress.  As one of the speakers in that session, I would
> > appreciate the
> > feedback.
> >
> > My paper was meant to start people thinking in the direction
> > that topic maps
> > might be able to be used to interchange knowledge bases.  I
> > also wanted to
> > demonstrate that it was, in fact, possible to build (simple)
> > knowledge bases
> > from topic map information.  I would love the opportunity to
> > work with the
> > knowledge engineering community so investigate this further.
> >
> > Eric
> >
> > <!--
> ****************************************************************
> > Eric Freese                                    Email:
> eric@isogen.com
> > Director - Professional Services - Midwest     Voice:
> 651 636 9180
> > ISOGEN International/DataChannel               Fax:
> 651 636 9191
> > 1611 West County Road B - Suite 204            WWW:
> www.isogen.com
> > St. Paul, MN 55113
> www.datachannel.com
> >
> ***************************************************************** -->
> >
> >
> > > -----Original Message-----
> > > From: Wrightson, Ann [mailto:Ann.Wrightson@sweetandmaxwell.co.uk]
> > > Sent: Wednesday, August 23, 2000 1:09 AM
> > > To: 'xtm-wg@egroups.com'
> > > Subject: [xtm-wg] Knowledge management claims re XTM (and
> > Topic Maps...)
> > >
> > >
> > > Hi folks...
> > >
> > > I don't know if you discussed this at all in Montreal,
> but I am very
> > > concerned that XTM should encourage/facilitate good
> working-with the
> > > long-standing academic knowledge-engineering community by
> > > a) being (very) moderate and realistic in its claims for XTM *by
> > > itself* as
> > > a knowledge modelling technique - XTM's main strength could
> > well be in
> > > providing the "hooks" which enable the wide range of
> > established knowledge
> > > modelling techniques to link to Web resources, and also in
> > > providing simple
> > > exchangeable *static* representations of (some) such models;
> > > b) working with folks who have done eg semantic modelling for
> > > years, rather
> > > than trying to (re)invent ways of doing it without the
> > benefit of that
> > > experience.
> > >
> > > This concern is based on back-of-room conversations at XML
> > Europe, where I
> > > found myself trying to bring some experienced knowledge engineers
> > > "back down
> > > off the ceiling", and into a more realistically
> > appreciative frame of mind
> > > regarding Topic Maps, after some rash claims from the podium...
> > >
> > > Cheers
> > >
> > > Ann W.
> > >
> > >
> > >
> > > To Post a message, send it to:   xtm-wg@eGroups.com
> > >
> > > To Unsubscribe, send a blank message to:
> > xtm-wg-unsubscribe@eGroups.com
> > >
> > >
> >
> >
> > --------------------------------------------------------------
> > ------<e|-
> > Need EDA tools on a short term or peak load basis?
> > Take a free 7 day trial!
> > http://click.egroups.com/1/8464/4/_/337252/_/967041084/
> > --------------------------------------------------------------
> > ------|e>-
> >
> > To Post a message, send it to:   xtm-wg@eGroups.com
> >
> > To Unsubscribe, send a blank message to:
> > xtm-wg-unsubscribe@eGroups.com
> >
>
>
> To Post a message, send it to:   xtm-wg@eGroups.com
>
> To Unsubscribe, send a blank message to:
> xtm-wg-unsubscribe@eGroups.com
>
>
> -------------------------- eGroups Sponsor
> -------------------------~-~>
> The Five Secrets Of A Successful Product Launch
> FREE right now at:
> http://click.egroups.com/1/8592/4/_/337252/_/967113859/
> --------------------------------------------------------------
> -------_->
>
> To Post a message, send it to:   xtm-wg@eGroups.com
>
> To Unsubscribe, send a blank message to:
> xtm-wg-unsubscribe@eGroups.com
>


To Post a message, send it to:   xtm-wg@eGroups.com

To Unsubscribe, send a blank message to: xtm-wg-unsubscribe@eGroups.com


-------------------------- eGroups Sponsor -------------------------~-~>
Thousands of Great Jobs, One Great Location!
Austinatwork.com. Great Jobs, Great Life!
http://click.egroups.com/1/7847/4/_/337252/_/967120593/
---------------------------------------------------------------------_->

To Post a message, send it to:   xtm-wg@eGroups.com

To Unsubscribe, send a blank message to: xtm-wg-unsubscribe@eGroups.com