Wednesday, November 23, 2005
Center of Excellence Proposal
à
ONTAC stands for Ontology and Taxonomy Coordinating Working Group
It is a working group of
Semantic Interoperability Community of Practice (SICoP)
Discussion Subject: Re: [ontac-forum] Sowa's Collection of Modules
Communication to BCNGroup regarding the ONTAC discussion from Ken Ewell
Hello Paul,
John Sowa's remarks below really illuminate the dichotomy between where everyone (data engineers) are focusing and where Readware focuses. It really helps me, for the very first time, to point out where Readware does something entirely different than others AND WHY it is probably a much better approach for obvious reasons.
John points out the problem of N-squared communications paths and he shows that the focus is on the external output and even alignment of the external output between entities. We (Tom and I) have always thought this to be a wrong-headed approach. It is not natural. It is not the way we think. Vocabulary is not the only issue, but also the internal/external syndrome so perfectly described by John Sowa. I hope you don't mind my use of the term syndrome there. I am suggesting that "best practice" is behavioral, pathological and destructive and I think John Sowa has given us sound advise why that remains an issue and is a major problem blocking widespread adoption of any approach to ontology or semantics.
Partly for the reason John Sowa points out: If that is the way it is done, large institutions, commercial or governmental, could dictate all the criteria for communicating with them, including vocabulary to use, on the authority of their accumulated corporate or monetary power. Who is to stop them; how many people know better and can do something about it? Not a bright future there.
On the other hand if you assume to begin with that there are an infinite number of communications events between entities of infinite types, you can build a set of facts on that assumption too. There is something shared by all interactive communicative events that can be captured and depended upon, irrespective of the type of entities that are communicating and the 'content' of the communication. Besides those concrete conditions, there are internal/external conditions that have to do with interpreting the content of a message in a communications channel. You know them as boundary and engagement conditions.
Readware works by recognizing these conditions and uses them to compute derived relations between external and interpreted forms of content processed by Readware algorithms. It is done by using an internal representation of concepts that (rather concretely) render reality. Readware algorithms derive directly from Tom Adi's original theory of semantics.
In order for your readers to to see how this approach works. I invite them to see the Readware Desktop GUI that that will let them access Readware functions through various interfaces. This is a package for information system developers that has not yet been released. It remains unfinished because a) I am not a programmer and have limited abilities, and; b) there is a general lack of resources at the moment.
Unfortunately this is ONLY for Windows platforms as it uses the Windows Scripting Host platform for desktop scripting in a way like an 'elephant language' is used for scripting complex interactions using an API to set of underlying functions. The underlying functions come from the Readware Ip Services that are installed and started on a local system. The underlying functions run anywhere Sun, Linux, Hp-Ux, etc., but the desktop device is just for Windows. User's can meta search Internet search engines, collect the links with a harvester or spider and compile them into their own searchable, shareable collections. User's can experiment with Readware topical analysis and clustering and with the document or file classification techniques.
So that I can converse with those trying the desktop app, whomever wants it should email me for a password to the installation program. The Windows desktop application can be read about and retrieved at the following address: http://www.readware.com/rwipgui/rwupgulrelnote.html
-Ken