OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

topicmaps-comment message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]


Subject: [topicmaps-comment] Everchanging subjects [ Notions haveexistence ...]


Hi Thomas

"David, do you see emergence as related to complexity in a topic map (or a
set of them), or do you see it as coming out of a simple system, like
cellular automata?  And if the latter, how do you think that topic maps
could fit in?"

I do not see a data structure by itself capable of emergence. At the present 
time topic maps are data structures.
and an "external" "engine" is used to do things with them. That is the 
external engine reads them and creates them, a sloppy metaphor is to say 
that the external engine "powers them" or ""makes them alive"". (gurk)

The engines it seems to me are programmed in such a way as to adhere to or 
implement the XTM spec. (well in the ideal world the engines do that). I 
think perhaps that the current engines are not graph processing aware and 
are more like the steam shovel doing good work only because a skilled 
operator knows how and when to pull the right levers. (ie activate the right 
internal routine))  [I fully expect indignation from engine writers. It is 
not my intention to criticise current engines. What I intend is to point out 
that the processing done by current engine designs is not situated, nothing 
more. As Thomas has pointed out there are a myriad applications where 
"business-programming-headed software" is just fine thankyou. (David agrees. 
generally))


I think there will be an eventual trend towards having a kind of 
graph-structure virtual-machine. these gsvm will be able to "process" or 
"operate" graph structures representing Topic Maps, RDF, DAML-OIL, 
Conceptual-Graphs, XML technology data structures, graphics graphs(SVG for 
rexample, maybe VISIO), and so on. Sort of how that 'you know what 
programming language' has a virtual-machine (but it is not a graph structure 
oriented processor per se).)

TMs  They are a graph structure.
TMs are not  (yet) general graph structures, in my opinion, they are 
specialized graph structures. Where they are not general graph structures 
they are not capable of "supporting" certain things. This is why some of the 
graph structure based operations some people would like to try doing arent 
possible right now. (I expect more criticism on that.) I do not see how 
anyone could say that TMs are universal turing machines for example. If they 
are Id be delighted to be given refernces as to where I can study how to use 
them that way.

By having a graph-structure-processing-aware ""engine"" or (software) 
processor "running" TM data structures , in the future it will be possible 
to incorporate/disincorporate other graph structure represented data such as 
RDF, or DAML-OIL , or CYC. Metaphorically speaking these different graph 
structures (at the present time) are like the different gauges of railroad 
track that have been layed around the world. When trains came to a border 
everybody got out walked across the border andgot on the other train. That's 
because the tracks were different gauges and a given train could run on only 
one gauge of track.

Also these GSVMs could have subsystems which use techologies such as
genetic algorthims, neural networks, fuzzy/rough sets,; which in effect 
allow the engines to learn the best construction methods and patterns  as 
time passed (learn by doing and observing). This means that with such 
subsystems that the engines quite potentially could become more efficient 
with the passage of time/"experience"/"learning". Genetic algorithms for 
example can allow an engine to examine a number of different structural 
patterns and ""figure out"" (""evolve"") which is the "best" particular 
structure for use. (If this sounds like an application of David Dodds' 
notion of "situated context" you're right).

cheers
David Dodds

PS: Paul Pruett may claim I am wrong about "situated context" being possible 
outside of special creation fleshbots (humans) but I believe
I will demonstrate that he is talking through his hat if he attacks me 
there. "Situated context" is merely one of the possibilities of *agency*, 
and frankly only warm pink fuzzy humanists still believe that agency is 
possible only in a human mind. That is so dreadfully anthropocentric as to 
be medieval.


>From: "Thomas B. Passin" <tpassin@home.com>
>To: topicmaps-comment@lists.oasis-open.org
>Subject: Re: [topicmaps-comment] Everchanging subjects [ Notions 
>haveexistence ...]
>Date: Fri, 19 Oct 2001 21:37:31 -0400
>
>[David Dodds]
>
> > Thomas said
> > "Do you think that machine processing will be/should be/has to be
> > very different in its essence from human use of the same topic maps?"
> >
> > I think of necessity that human *use* of a _given data structure_, topic
> > maps in this case, will be more sophisticated than the *use* of the same
>ds
> > by a computer. This is true because (except for possibly severely 
>retarded
> > people) most people bring the sophistication of situated activity to 
>bear
>on
> > use of TMs , and they would just as much if they were shovelling dirt
> > instead. Humans are INTEGRATED in the use of their mental powers, they 
>are
> > "conceptually organic" as it were, they use ganzfeld, they make/use
>context.
>
>[...]
>
> > the difference between agents (such as humans or programs) doing organic
> > perception and agents doing inorganic perception is major, it is the
> > difference between what a person brings to bear using a topic map and 
>what
> > nematode brings to the same topic map "usage". Only a pointy headed
> > programming-only immersed person cant see things like tacitness and
> > metaphorical/analogical activity accompany human activity and just aint
> > there at all in business-programming-headed software.
> >
>
>Well, there is a definite place for the current type of
>"business-programming-headed-software", isn't there?  After all, we 
>probably
>don't want TurboTax coming up with unexpected emergent behavior at tax 
>time.
>It's just that we are wanting more from our systems now.
>
>David, do you see emergence as related to complexity in a topic map (or a
>set of them), or do you see it as coming out of a simple system, like
>cellular automata?  And if the latter, how do you think that topic maps
>could fit in?
>
>
>----------------------------------------------------------------
>To subscribe or unsubscribe from this elist use the subscription
>manager: <http://lists.oasis-open.org/ob/adm.pl>


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



[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]


Powered by eList eXpress LLC