Saturday, November 06, 2004
New discussion about data regularity in context
and structured ontology
On the Nature of science and
cultural Polemics à
On the discussion of data
regularity à
On the Stephenson Cyber Attack
Taxonomy à
The development of some preference ordering and rules, between substructural ontology and functional/systems ontology must reflect some natural phenomenon in some way. Describing this phenomenon will be the challenge to system theorists.
Part II, Adi structured
ontology
A first step is to make the types of observations that you have made. The justification for the observations rely on your own confidence that these observations are revealing and on a sense of agreement that I, or someone else, might have about the use of the three processes:
{ assignment, manifestation, containment }
and about other language used in Part I and Part II.
Your work makes observations that fit within a method that brings sense and satisfaction to you. The sense of satisfaction can be shared. There are many elements that cohere together to produce something with completeness and having a consistency that we need in our software.
The objective test is in how well these cross scale rules work in terms of setting up a cultural ontology, this is Part III, and in identifying the communicative interests of people using a language that Readware has instrumented, for measurement of the substructure.
Finding a comfort level to the expression of objective knowledge is one of the challenges that we, the core team, are facing. In this regard I have two comments, (1) objective outcome metrics and (2) the generative methodology. (On the new work about cA. eC, gFT and the generative methodology à [28] )
Objective
outcome metrics
The precision / recall metrics demonstrated by Readware in the past is impressive, and yet the metric is one that measures this notion of retrieval. The next step is to show that a real time representation of many textual elements, from a community, can be summarized into a poll type representation of the social discourse in a historical context.
That some point, we will have a system that measures the social discourse at a substructural level and then allows differential computations so that variations in how individuals and communities experience their knowledge is taken into account. Ken and I have been talking about how to establish a market for social discourse measurement capability. The development of literature repositories and content management systems has stalled because most business processes do not have a background in the preliminarily elements of the knowledge sciences. (For more on the cultural and educational background needed, see Section 2 of the Preface, Knowledge Foundations).
The Readware software already has a high degree of control over specific parts of a conceptual roll-up over a text collection. But we need a tutorial on exactly what user control exists now, and we need to plan ahead as we develop additional user control.
What is not under the control of the user? Those things like the preference rules and the process delineation:
{ assignment, manifestation, containment }
would, I imagine, be established as self evident and thus not subject to modification. Such self-evidence is properly called an ontological commitment. The ontological commitment follow an enumeration of a dimension, the process dimension, within the complex framework that Adi has presented to us.
The
generative methodology
As a principle, the ontological commitments are independent on a generative methodology that produces new structured ontology based on a descriptive enumeration of some number of dimensions. The Adi structural ontology being described in Part I – Part III can be seen as a specific application of a methodology. The reasonableness of the ontological commitments is a key to an evaluation of the Adi ontology as a theory of language. It is, however, the combination of the justification for a formal construction based on an induction, or set of inductions, and experimental work.
The control over a structured ontology, like the Readware ontology, can be presented more completely if the ontological commitments are revealed completely, as Tom Adi is in the process of doing.