Friday, January 27, 2006
[148] ß [parallel discussion on
generative methodology (Judith Rosen)
[147] ß [parallel discussion on
generative methodology (Peter Krieg)
[368] ß [comment on four issues
(Richard Ballard)
Four Issues about Ontological
Modeling
Communication from Paul Werbos à [367]
Communication from Paul Prueitt
The conversation needs to get complex at this point, and I am not sure how to proceed. Dick Ballard has developed a history of his work on n-ary information processors ([368-1]). In this history he gives knowledge of simple data coding techniques that make the new generation of n-ary processors several orders of efficiencies higher than the relational database. Ballard and I have often talked about n-ary encoding and information theory in the various beads, of the glass bead games.
The Hilbert engine (by Bjorn Gruenwald) approaches the same data encoding but without an information theory. Gruenwald's work is explained a number of times in my research notes and in the bead games. A google on "Keyless hash table" gets "with quotes" only five pages, all written by myself.
When Dick Ballard talks about the planned limitations of Oracle and Rational databases (etc)
"From the perspective of triples. the database starts to break and
Oracle and
Rational with their UML language invented modeling rules to never let
modelers ever see or try this'
he is really talking about a business strategy to keep market share. The result, as Dick points out, severely limits how computers are used as data management systems. BCNGroup claims of “false promises” maintained by the ONTAC working group is based on evidence that business strategies have institutionalized as federal government IT procurement rules.
The move to semantic web, XML, is a partial revolution; but XML suffers by not knowing about key-less hash tables, and the SW suffers from an improper understanding of the nature of formal logic (as has been discussed in the bead games). So we take two steps forward and one step back. The two steps forward are exciting and offer advances over the relational database and data models by providing well thought out information models such as the OASIS SOA IM (Service Oriented Architecture Information Model).
So I will let the others read his note, with his permission; reminding us that the focus of this discussion is about the formative importance of how natural systems interact with each other (ie their relationalism - to use Krieg's term) . { since "relational" is already owned". }
The use of language debate that we have had on an off is also relevant, since "like "complexity" which is not complex, and "expressive logic" which is not expressive; relational databases are not fully relational. So developing a branding language for the Second School is essential.
{My current effort is in developing SOA (Service Oriented Architecture) methodology that leads to a business environment where economic value comes from having the most efficient information theory, and encoding technology, available within an interacting human community. We are at a new point in history where the business strategy of the technology provider does not trump the needs of the market for true data and process interoperability. But branding language is needed that quickly shows what the problems are with OWL and more generally knowledge engineering.}
Many key terms are defined, largely by IT business strategies, to close off an understanding, or path to understanding, that is not "owned" by a business.
Here in lies the challenge for knowledge scientists. Our strategy is to create a "Second School of Semantic Science" wiki, where the Second School can allows the First School to define its terms, while the Second School ALSO gets to define its terms, and the two viewpoints are side by side.