[topicmaps-comment] situated context and topic map generation

[David Dodds <drdodds42@hotmail.com>]

Anyone in the topic maps domain who is interested in the progression of my 
work in the areas of metaphor/analogy processing and (the related areas of) 
"situated context processing" is invited to read my postings at kmci virtual 
chapter , it is a yahoo groups. I have been posting a series there, with the 
subject heading of metaphor/ analogy, and also situated context, 
occaisionally different headings. The series has been underway there for 
many weeks and there has been keen interest in it from some of the readers.

because it isnt always easy to locate all of things on various elists, such 
as the one I just mentioned, I am in the process of gathering all of the 
content of those postings together into one web access point. In several 
days I will be posting here a URL link to this place where the series todate 
is collected. The series is ongoing at this time, throught e auspices of 
kmci vc, I invite any interested readers from this forum to have a look 
there.

Amongst the things that I have already posted in the series is a set of SVG 
based animations which demonstrate "computational" instantiation of notions 
from Edward De Bono's "management" concepts, which he illustrated in his 
book Atlas of management Thinking.

here are a few links to such illustrations from his book (at the end of this 
posting).

Both in the WROX book XML Meta Data and in the metaphor/analogy series I 
have posted I discuss how to capture and represent "knowledge", such as the 
semantic contents of a diagram. In the book it was an electronic circuit 
schematic, in the postings it was deBono diagrams, and one specificaly about 
concepts involved in group effort achievement and thwarting thereof.

I used topicmap in the book to illustrate how the diagram could be 
"understood", and I also am using topic maps in the series asa means of 
coupling TMs, RDF, XML SVG. I use the generation of topic map (serialized 
code form) as an example of results afforded by what I call situated context 
processing. Since computer code is involved anyone is free to see how it is 
done. There is no magic or religion. Von Neuman computers are just too 
"dull" at the present time to be able to "finesse" anything, the way humans 
can when the do mind-based processing sch as Gedanken experiemnts. I 
strongly believe that "mental-finessing" will become a very imporatnt topic 
(no pun intended) in the forseeable future. The "problem of finessing" 
occurs when human-mind-based system design/construction is deemed to be 
"whole" and "complete"
(coverage complete), works when the system is "run in the human head" but 
shows one or more nontrivial oops when run on a computer.

The detection and handling of (human mind) "finessed" oversights in designs 
of knowledge technology systems will necessarily become a legitimate area of 
both "computerdom" and "psychology/philosophy".
Some insights into the nature of know "finess"-based problems can be gleaned 
by reading "What Computers Still Cannot Do".

I think we are all going to have lot of fun. Some of the topic maps 
community it seems to me just wants to get a solid workable IT type topic 
maps nailed down so that we (the topic maps community) can get the 
public/potential users to use these things/buy them as a techonolgy.

I think one of things that needs to be done is to make sure they are not 
perceived as some new mysterious incantation such as many see RDF.
In the public's eye RDF has not been yet explained clearly enough for the 
mto get it. We TMers must explain it to them so that they get it. perhaps 
some exciting TM applications is what is needed?

IT type topic maps must survive and flourish, in my opinion, otherwise the 
great unwashed public may see TM as Beta tape and turn to some version of 
VHS! (gurk)

As IT type TM survives and flourishes (spreads) we are then in a position to 
make effort into what I might call an IT+ type of map, one which is a 
graph-structure based reprsentation which uses an engine which is 
""cognizant"" of graph structures like RDF, DAML-OIL, CYC as well as XTM. 
"Interoperability" will bring further acceptance of TM to those consumer 
folk out there. They probably wont spend money on simply the beauty of the 
design of XTM.

My postings (at KMCI) will continue to explain/discuss how I go about making 
code for situated context processing and metaphor/analogy processing.



Thomas said"
1) What can topic maps be used for, and can they include sufficient
knowledge representation that maps can be used in the creation of situated 
or aware systems?"

David Dodds"
a major part of the point (and content) of my postings/series on 
metaphor/analogy and situated context processing (so far at KMCI)
has been to show code and explanation as to how to have/create situated 
context systems, of which situated generated topic maps are an example of 
output/results of the process. (as well as SVG visual reification)

I also started to explain what I called "aware1" early this year (again at 
KMCI). "Aware1" is the notion of awareness-like processing (in a non-brain). 
The digit one on the end of the word aware is to signify that this 
"awareness" is not intented to be confused with the "thing" that many people 
mean by "HUMAN AWARENESS". Paul and others will understand, by virtue of the 
1 on the end of the word aware, that aware1 is not about how human mind is 
aware, it is about _A_ way (as opposed to _the_ way, such as flesh way) that 
*agency* can be instantiated. I see that I should return to further 
elaborating aware1 in the KMCI discussions which I post."


Thomas said"
My reading is that David is pretty sure they can, Paul hopes that they can, 
and I hope so but I'm still unsure.  Am I close to the mark here?

David Dodds: "yes, in my opinion".

Thoams said"
2) How can useful topic maps be created - especially as a group or social 
process - and should we expect them to be relatively static or constantly 
changing and evolving?'

David Dodds"
I think, as I said earlier, we need to separate the two kinds oftopic maps. 
The IT type topic maps must survive (out in public). Some of thetopicmap 
community firmly believ that only static maps will work out there. maybe so. 
Ifthat what it takes to get the "crowd out there" to get using them then let 
it be so.
I think as an advanced knowledge representation system that the feature set 
will have to be added to/expanded. But these additions may be unfavorable in 
the eyes of part of the TM defining/designing community.
Perhaps it will be necessary to have two or more kinds of TMs in the future. 
Those that adhere to XTM X.Y spec and those which have some
(from XTM spec point ofview) "nonstandard" parts to them. I think the buyin 
of these second kind of TMs will critically require the use of programs 
which can perform an "automatic transmission" kind of behaviour, allowing 
"dynamic additions" to the XTM spec. Doing the automatic transmission in a 
stable way may become a focus in the XTM engine building world. Again, I 
think that a graph-processing-aware engine would help this scenario alot.

David Dodds

de Bono   illustrations
http://www.sixhats.com/images/atl_013.gif
http://www.sixhats.com/images/atl_014.gif
http://www.sixhats.com/images/atl_015.gif
http://www.sixhats.com/images/atl_016.gif



_________________________________________________________________
Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp