topicmaps-comment message
[Date Prev]
| [Thread Prev]
| [Thread Next]
| [Date Next]
--
[Date Index]
| [Thread Index]
| [Elist Home]
Subject: [topicmaps-comment] Letters on Mark X Series Editing Tools &Experience -- An Invitation and comment
- From: paul <beadmaster@ontologystream.com>
- To: Topicmaps-Comments <topicmaps-comment@lists.oasis-open.org>
- Date: Sat, 08 Jun 2002 09:39:49 -0400
Comments from the Topic map
community?
(Open communication,
please circulate)
Dick
Ballard
Knowledge
Foundations Inc.
Dick,
The intent of your
new letters, to the categoricalAbstraction forum, is perfect, in that the
barriers that the community has in expressing intellectual property is quite
profound. Your willingness to actually talk about how the Mark 2 works is
important to the community, and unlike - for example - the Cyc ontologies where
there is no ability for those who could make an objective evaluation to make
this evaluation. And of course, the topic maps ontologies seems to never
really occur in a way that is visible either. Cognitive graphs (Sowa's
work) are the same way, in that after working through the technical literature
he has written one is left with a great need to see something work in real
settings.
The philosophical
grounds for understanding why machine ontologies, up to now, do not lead to real
working systems is made from an examination of the illusionary nature of formal
systems.
The principles that
are being designed into the Mark 3 address the required openness in the
formalism and to human interaction in a way that is consistent with what the
Russian logicians wanted from QAT and semiotic system for control of
(whatever):
and to the work of
Peter Kugler and others.
As we talk about the
Mark 2 and the Mark 3, there will be insights both by yourself and by
others. These insights are the substance of patents if handled in a proper
fashion. As I do think that the SAIC/OntologyStream proposal to NIMA
will be funded, we might wish to put in place a process to allow a restricted
community an ability to quickly disclose as patents those insights that none of
us have had as yet. I call this an IP Glass Bead Game.
There are currently
four such IP Glass Bead Games being played. I happen to be the Magister
Ludi in each of these, except one where Don Mitchell is playing this role.
I suspect that there will be many hundreds of these games being played as
we start to encode knowledge technology innovations into a special CoreTalk
operating system and memory manager (which can only now be discussed under
non-disclosure agreements). Again this notion, of community building
through the capture of IP ownership, is the NIST proposal, but we may find
venture capital even if NIST can not quite come to break down this barrier as
disclosed in our Three Year Plan:
Someone somewhere
has got to express the science of knowledge systems in a clear fashion without
all of the subterfuge that we find in the XML and ontologies communities.
For example, the only real disclosure of what individual innovators are doing is
made into a patent process that has a 3 and sometimes 4 year period where the
disclosure is not public - while the PTO makes an evaluation. This is
useless in the context of establishing a new discipline. How can one
innovate on secret "disclosures"?
The PTO is broken in
many ways. Over time, the PTO's behavior may develop to actually be
un-Constitutional as well as being simply confused about the nature of
information and knowledge technology.
The hard issues in
knowledge representation is made only harder in a climate where some workers
have only partial glimpses of a complete knowledge system. It is often the
case that these individuals present themselves as having seen the holy grail -
because it is required by the management and investment airheads that
anyone asking for investment make these types of ridiculous presentations.
This investment behavior badly screws up everything that needs to be done to
develop a discipline and a community. The DoD cottage industry managers
finish the "innovation road kill" phenomenon by disallowing any real innovation
into the government... as we all see with the fact that at the FBI that the
notion of full text retrieval is a dream that might take 2 more years to
implement. Fuzzy search is specifically discounted, and the notion of
concept search is treated as an embarrassment. One has to ask, as
some Senators are asking, why the DoD cottage industries are acting as gate
keeper to keep normal and mature COTS based capabilities away form the service
of our nation. The Senators should ask the scientists, like me, who
have been consistently blocked from deploying full text and conceptual search
technologies for the past decade.
(Perhaps the cottage
industries believe that the American People have not yet paid enough for their
high priced gate keeping. I personally think that we have legal and social
crimes being committed by those whose integrity we are asked not to challenge,
and I think that this will continue until Congressional hearings reveal the
complete and total truth about what these IT companies are doing to "earn"
a profit by institutionalizing these ugly problems. I specifically have
much to tell in regards to my own experience and a set of other case studies of
similar experience by other scientists.
A solution to the
cyber terrorism problem is available, not only in my work but in the work of a
number of under funded and abused start-ups.
This solution is
similar to the solution to the cyber porn problem, in that the system preserves
the problem because it is a good problem "for the system" - in spite of what the
problem does to human culture.)
And ultimately we
see that the IT investments themselves served the ego of the investors initially
and then is spoiled because the IT product is not correct. So we see that
there is no ROI with the current investment behaviors. The system
tolerates this because of the (often undeserved) concentration of wealth that
occurs under our form of capitalism. So the wealthy can afford to make
deep and profound mistakes in what is funded. The loser here is the world's
social structure.
Out of a Darwinian
process of random selection, comes a few survivors.
The advantage that
the Mark X series has is the historical presence. There is a history
here. Another advantage is the information theory that you developed and
vetted with important theorists, and for which much more can be said than has
been said - or perhaps even imagined even by you or I.
Dick, and the others
in this forum. I wish to address this issue of a IP and proposed
scholarship and patent Mark 3 - based knowledge base.
Fist, because it is
important to the community; this economic power of a common repository of patent
and scholarship disclosures.
Like the moves (or
books) where evil is defeated only by turning the evil power into itself;
capitalism's reductionist reinforcement has to be used to protect the new
knowledge science and technology. This is the brilliance of the BCNGroup
Charter, in that a mechanism is envisioned to take property ownership as gifts
form creative people who do not care to own anything, and confers this into a
community ownership in the form of Patent and Trademark Office approved
patents. The community ownership is then dispersed, completed, (this is to
say purged) from the BCNGroup by a scientific committee making gifts to
individual scientist or innovators. The purging of the money must occur
once a year, by Charter. This is done through prizes and university
scholarships.
I have a
relationship to M-CAM that allows me to run patents that are submitted for
review and potential inclusion through an automatic intellectual property space
mapping technology. This service general cost around $15,000 from M-CAM,
but the BCNGroup has the ability to format patents in a way that allows M-CAM to
batch process 5 - 9 patents in a cluster for around this price. As I know
the internal architecture, this work is streamlined.
What is then
possible is that an innovator or group of innovators, such as a group of people
working under your direction on the Mark 3, would be observed by a Magister
Ludi
or group of
"advisors". The patent disclosure is homoeomorphism to the PhD thesis in
that:
1) broad competency
must be demonstrated
2) an original
contribution to human knowledge must be made
2) a literature
review must specify the relationship of this original contribution to the
elements of the literature review.
Failure to know of
someone's work that is similar to one's own is to lose the opportunity for a
patent of a PhD. Many many individuals have forgotten this simple
rule.
So the BCNGroup
wants to assist in the proper disclosure of IP, first in the form of patents and
then later on in the form of virtual PhD processes.
The Mark 3 is the
proper knowledge system to encode the community knowledge of knowledge science
and technology.
One is constantly
dealing with protecting one's IP in a situation where the immune system of the
status quo does not want to allow knowledge science and technology a place to
root. However, the BCNGroup strategy is to use hundreds of commonly owned
patents on processes that are valuable when rendered as software, to root
knowledge technology in a fertile valley, rather than on some rock ledge.
The time is right
for this process to start.
Please
consider joining the BCNGroup membership.
Friends:
Last February, I talked of
beginning a new series of letters on the Mark X
EditForm, Builder, and the
User Editors. That was before the Knowledge
Technology Conference in March.
After that conference, I appreciated better
the gaps in the technological
community's knowledge of knowledge. So I
embarked on a series of 7 letters,
that surveyed what I thought were the
most important "missing pieces".
Wojtek and Jack promise not to let me off
the hook in answering additional
questions with respect to those letters, so
consider that discussion as
still open.
This intended new series is completely at the other end of
the spectrum. It
aims only to write the content of a knowledge model into a
simple ASCII text
file. With that start, it will follow each editorial step
needed to prepare
these files for a semi-automatic process of parsing and
mark-up aimed at
integrating this model with thousands to hundreds of
thousands of other
models. This process of editing is at the center of any
large scale
knowledge capture project.
Over the past 16 years this
process has evolved through two distinct
generations (Mark 1 & 2). My
aim is to describe and deconstruct the Mark 2
process and all the key
issues of editorial thinking, examining closely each
distinct tasks and its
purpose. Then "brainstorm" each goal and process in
search of better, more
productive methods. Ultimately I will try to fashion
a tool design that
makes such best practices the obvious thing to do when
used by that rather
ordinary sort of worker (no disrespect intended), drawn
to editorial
professions. In the space of several months I plan to refashion
the
editorial process, its work products, and then sketch the tool designs
in
preparation for programming and prototype development.
What is striking
I think is that when we watch Knowledge workers, they
appear to be doing
things we have all done before. But when we put a
microscope to their
thinking and intent as they work, we see levels of
detail and rigorous
concerns that are many levels deeper than most people
have ever thought to
think before. To be useful to real world editors this
work must have the
checklists, the tests, the clear decision criteria that
goes directly to
some reproducible way of making every abstract idea utterly
and
unmistakably concrete. As with children, every instruction that is found
to
stop short , leads immediately to the constantly repeated question,
"Yes,
but WHY?" Finding yourself explaining "why", is always a voyage
filled with
self-discovery and constructive re-creation.
THE TOOLS
IN QUESTION:
EditForm -- is an ASCII text file that is read and parsed
by the BUILDER
tool. In production, everything going into the Knowledge
Layer goes in
through an EditForm. In Mark 2, if changes are to be made to
the Knowledge
Layer, the editor constructed an EditForm to identify the
concepts to be
altered and then asks the BUILDER to markup and "expand"
these targeted
concepts. The Builder performs this mark up and returns the
editor's
EditForm to show every possible connection to other content items.
The
editor makes a copy of this "as is" EditForm, then constructs their own
"as
should be" EditForm version. If their changes are "authorized", the "as
is"
form is marked as a deletion and the "as should be" as the "new" body
of
connections constructed in its place.
There are a great many new
issues in the transformation of Mark 2 EditForms
to Mark 3. The most
important of these is the basic Mark 3 grammar and all
of the ontological
assumptions and conventions it entails. This grammar is
the signature
defining every representational form that Mark 3 can
incorporate within a
knowledge base. For everyone interested in the
formulation of a system of
knowledge representation beyond Mark 2, Topic
Maps, and all the rest --
this discussion should fix the final form of Mark
3.0. You will want to be
in on that debate and its consideration of all
possible issues.
This
is clearly the place to bring up your own lessons learned from other
tools
and ontological concerns.
BUILDER -- is the production tool that builds
and maintains a single
knowledge layer over time. Versions of this layer
can be "published" from
time to time as a proprietary knowledge asset. In
this sense we speak of the
BUILDER as a printing press or production
station. Mark 2's builder was site
(or seat) licensed as a commercial,
small business enterprise tool from
1991-1999 at $50K per 2-seat license
"as is" or for upwards to $100K for
licenses that required substantial new
capabilities to be added and
delivered. Each license provided 1 year of
updates, hot-line support, while
setting no production limit or royalties
on the products it produced. In
most cases Mark 2 licenses were sold to
project sponsors, who ultimately
would hold copyright and title to the
knowledge asset under development. By
owning the tool license and the
product, they could alternately internalize
the production process or
outsource it to yet another service provider.
Most often the initial
development services provider was a team of two or
more companies. One
company, the prime, was a system integrator or
consulting firm with a long
and close relationship to the project sponsor.
The other was a unit from
Knowledge Research, my "project-oriented
production shop", an especially
organized group of experienced and highly
skilled professionals capable of
monumental feats of production, of cloning
assistants, and of training
aides to quickly grow their total labor force to
match projects of any
size.
The Mark 3 BUILDER is an evolutionary production system that will
appear in
versions suited to a number of markets -- teaching, professional,
small
business asset production, and large enterprise development and
support.
This discussion will focus first on the small business knowledge
asset
production shop, whose expected annual business should exceed
$1
million/year. This level of business can support one or more full
time
production teams.
The key design goal of this BUILDER version
is a productive, collaborative,
and profitable process of knowledge capture
and work flow. Its design as a
process management, work scheduling,
ontology development, and collaborative
feedback tool maximizes our
opportunities to embed all our best practice
work experience into the tool
itself.
USER EDITOR -- is a highly interactive editor intended for
individual
students, professionals, and knowledge workers who are creating
and editing
their own original knowledge assets in an ordinary office
desktop or laptop
work environment. This environment is less collaborative
and less focused on
high production throughput and large scale asset
integration. This tool will
receive greater attention for Mark
3.1.
JOINING THE DISCUSSION
We can broadcast our discussions on
the open forums, but I expect there to
be a variety of sample documents and
intermediate work products exchanged. I
am hoping that those seriously
involved in this discussion will keep us
aware of their interest and
constant attention, especially by contributing
regularly. As we go I will
build and maintain a distribution list of
frequent contributors and be
sending my letter attachments directly to them.
Those sharing attached
artifacts may want to build similar lists to
facilitate their own direct
distributions to this "inner group".
Please respond directly, if you
aim to be a regular participant and preview
for me some of the issues and
discussions most important to you. I will try
to put together some
checklist or rough schedule and benchmark about where I
think we are -- as
we go along. Let me know when you would like to jump in
with some major
initiative or teaching and discussion topics and I can make
way and give
your ideas full focus.
Give me a "heads up" on your interests this week
and I will put up a first
outline and introductory starting letter by next
weekend.
Dick
[Date Prev]
| [Thread Prev]
| [Thread Next]
| [Date Next]
--
[Date Index]
| [Thread Index]
| [Elist Home]
Powered by eList eXpress LLC