[XML-DEV Mailing List Archive Home]
[By Thread]
[By Date]
RE: [xml-dev] Invitation to metadata dictionary wiki - meaningfue l.org
- To: 'Elliotte Harold' <elharo@...>
- Subject: RE: [xml-dev] Invitation to metadata dictionary wiki - meaningfue l.org
- From: "Bullard, Claude L (Len)" <len.bullard@...>
- Date: Mon, 30 Jan 2006 08:33:41 -0600
- Cc: xml-dev@...
http://www.msnbc.msn.com/id/11009379/from/RS.4/
And I ask again, crowds of what?
It is easy to construct a crowd where everyone is
talking past each other (few are listening or
actually logically reasoning). So crowd of what
and the subject matter are important. I suspect
that given some topics and some crowds, it can
be very hard to get more than a clear picture of
the polities and the memberships and that is
created by a different set of analysts/crowd.
Wikipedia and any other documented discussion
has a feedback effect (that is the averaging
function) if serious participants stay current
and reanalyze past threads. There is a difference
in PageRank functions that order a search and
analysis functions that select the answer from
a candidate list. That is precisely why an
ontology is layered over a service (preselected
choice of choices). The connectedness (density
of connections) of the ontology now comes into
play and has operational or scaling effects based
on the criticality of a choice.
Quick and don't look: what is at the upper left
of the majority of menu toolbars? Would you be
affected if it randomly moves to the center or right?
len
From: Elliotte Harold [mailto:elharo@...]
One of the revelations of Wikipedia is that crowds are pretty effective
at self-selecting and self-policing. A random crowd would not be
effective at picking surgical tools. A self-selected group of people
interested enough in surgical tools to participate in a discussion of
surgical tools likely would be; and in fact probably would do better
than any one individual would, no matter well educated and credentialed.
|
