These discussions give a background to
Safe Net and National Project to Establish
the Knowledge Sciences Curriculum
The background is just laid out, so that
others can see where
the process is and how they might
contribute to it.
From L Van Warren :12/12/2003 6:37 PM
Topic:
Knowledge mapping ecosystems
proposal for NSF
Highlights:
- Chris suggested that
knowledge mapping is at the "tipping point".
- Think about building a
technology for scientific "smart mobs" (Google count 91,500) or
"flash mobs" (Google count 28,500). This is a super aggressive
form of ontology building for problems of criticality.
- A measure of
effectiveness is the time for a collaborative group to visualize and
simulate an ontological work unit.
- An effective team is
going to populated by specific expertise at several layers.
- Knowledge maps will
transcend spatial scale.
- Discovery and
understanding lend themselves to industrial revolution "assembly
line" metaphors. There is leverage.
Discussion
We reviewed:
- your three power point
proposal slides.
- knowledge mapping
demonstrations and including a
Flash animation.
- knowledge browser KnowledgeGAZER™
the first web based knowledge browser applied to complex disease.
We talked about:
- Separating 'what' from
'how'. What we know, versus, How we figured it out.
- The NPL parser that Daniel Sleator of CMU has
created. Could enable rapid progress in the journal data mining.
- Flight simulator
"levels of detail" as being a working analogy for what a user
should see.
- The work that Paul Pruett
has done in categorical abstraction. Paul's algorithms suggest that
"levels of detail" and compression of knowledge maps is a
next step. Paul is also working toward a K-12 education curriculum that
would include these ideas.
- The work of Kathleen Fisher at San
Diego State University. Kathleen is a long time advocate of idea
mapping for teaching biology and has released a desktop applications
called Semantica.
- The work of Mark Diggs
formerly of Ontology
works, an ambitious ontology software project.
- How ontology
construction is the first step that is qualifying objects and
relationships that exist.
- How relationships are
quantified, you used the mapmaker's term ordinated. This
deep implications in the knowledge mapping process.
- Back ending
quantitative knowledge maps with realistic simulation to close the
feedback loop.
- Pam talked about
the knowledge map to find out what the simulation told you – as an
end product, as well as a start product.
- The knowledge mapping
assembly line, to be populated by, domain experts, software engineers,
linguistic experts . I have a slide on this, but it really needs to be
nailed down as a methodology, for which specific recognized roles exist.
See my article on movie
production roles as an example. A similar document needs to be drafted
for knowledge mapping.
- The possibility of
convening a small knowledge mapping conference. One of the emerging areas
of this is intellectual property.
- Creating a web
interface for curated knowledge enabling Wikipedia-like
construction and use of maps. I have done some conceptual mock-ups of what
this might look like. I will add these to the knowledge mapping web page.
This could become a web portal, curated by someone like CERL or Argonne
perhaps.
- The necessity of exploiting
human perceptual characteristics such as eye hand, wiggly maps and color.
I am getting ready to post some advance in this area.
- Full blown
"people in VR suits interacting with 3 dimensional knowledge". I
prefer lime green for my VR suit. That way I can be invisible.
Three types of journal
data-mining.
- totally automatic –
done for my complex disease examples – good for speed, bad for
relationship identification
- semi
automatic – done for my cancer gene examples - ok for
speed, ok for relationship identification
- totally manual – bad
for speed, great for relationship identification.
Other Items and Efforts on the
Radar
Well that's it. Poof! I'm spent.
Cheers,
- Van
L. Van Warren
Director