[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]
Subject: RE: HM.Requirements: HumanProof
Identification processes use data (observable) to create candidate
lists. Then another more focused process is engaged which
narrows
and
verifies. The principle of view dimensionality involves the
processes that construct the view by selecting choices of
choices and choices. A report may be created by query
or
other means. The role of inference engines (eg, RDF systems)
may be
in guidance for each step of focusing the view, that is,
candidate selection. Most of the technology for query
refinement
comes
into play.
However, the data mining aspects involve:
1. Names and types of information that are in the database
(with
the obvious issues of namespaces). This is something
schemas do well.
2. Relationships that exist among names and types of
information (something Topic Maps and RDF do).
3. Policies regarding the collection and dissemination
for
the names and types of information.
4. Rules for follow on actions given identification and
classification of some type and an instance of that.
The
system deals with stereotypes until an individual
is
selected and at this point, the issues of identity
come
into play. However, there are multiple
systems
for
ascribing identity to a person and any system
of categorization and selection should work
with
all of these. Therefore, a web-centric system
based
on URIs is useful within a web system of
identification
but it is extremely cherry to assume it is in
use.
That is, yes a human could be considered a
resource,
but I doubt they see themselves like that and
part
of the issue of HumanML is that there is an
inner
schema(s) and an outer
schema(s) with properties
that affect both self-identification and
relationships.
The only resources we can count on having a
URI
are our own
models.
Relationships among individuals are
isolated and then
the
system looks for known patterns of behavior to
identify types of activity that may initiate further
actions. Visualization systems are very important
because the level of detail is high and in some
situations, rapid response requires the pattern
identification to be very efficient if somewhat gross.
Later,
it is refined for other uses.
Len http://www.mp3.com/LenBullard
|
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]
Powered by eList eXpress LLC