[humanmarkup] The Relationship of RDF to XML in HumanML

[Rex Brooks <rexb@starbourne.com>]

Hi Everyone,

This is the message I referred to earlier in my post on proxemic. 
Since I am still pretty darn excited about it must mean that either I 
am crazy, or there is something to it, owing to the fact that I have 
not been able to get back to this since Wednesday, and it was not a 
big part of our meeting, but it IS a big deal to me. Manos Batsis is 
going to be working on an RDF sample for us which we will hopefully 
see early next week.

The significance of this is that while we have been toiling along on 
the Primary Base Schema, and even considering dropping the division 
of Base and Secondary Schemata as well as looking at the very 
promising semiotic processor experiment, there has been a fairly 
important issue simmering on the back burner, which James Landrum has 
called to our attention a couple of times, and which the issues 
raised by the semiotic experiment bear on.

That issue is the various standards, and most particularly items, 
terminology, elements within those standards, whether they be 
promulgated by the IETF. ISO, W3C, Web 3D Consortium, OASIS, or 
HR-XML.org, with which we need to consonant, interoperable, and 
compatible, yet with which we will inevitably have differences.

James has been contributing resource references and I have been 
citing existing standards that bear directly on elements in the 
Primary Base Schema, and most recently we have considered what might 
seem like a very simple concept, measurement. As it turns out this is 
neither simple nor even very easy. In our meeting we tossed around 
some further concepts relating to how we can apply the element 
measurement_unit to subjective qualititative elements such as 
emotional intensities, which will still need to have relative 
measurements by which these intensity values can be used in a 
standardized way.

What Manos can provide, through his work on RDF is a way to put such 
subjective measurements into a common context by directing processors 
called on by application writers using HumanML to the resources we 
cite when we want to use a term with a qualitative, subjective 
measurement. Although this is not the particular example Manos is 
working on, what RDF can do for each of the areas where we may need 
to have slightly different or uniquely HumanML meanings for terms is 
to set up RDF schemata that tie the terms we want to use in a certain 
way to a resource document which gives that usage the meaning we 
want. I envision this as mostly a way to use resources such as James 
has been submitting to us for consideration and other academically or 
institutionally accepted and formalized usages for terms.

What I think we can do very effectively is to take the various 
enumerations or associations (used in the published topic maps sense 
and comparable to what Ranjeeth calls codelists) we wish and connect 
them to our elements. What we can do with the Secondary Base Schema 
is to compile these enumerations and place them into our huml 
namespace that will be maintained by OASIS once our specifications 
are approved as standards (What an optimist!) When the resources we 
want referred to for precise definitions of these eunmerated terms 
are so connected by the corresponding RDF Schema, we will have a very 
solid foundation on which our many applications can build in 
consonance with accepted science and technology.

At the upper level, to be thoroughly connected with Knowledge 
Management Science, we can have our entire language grounded in the 
DAML-OIL Ontologies. For some better understanding of that you can 
refer to the HM.frameworks document. However I don't think it is 
necessary to stipulate that now, nor is it necessary to specify 
DAML-OIL as opposed the Cyc. We can state that our work is based on 
an understanding of the importance of being aligned with the most 
accepted Ontological systems. At some point in the future I suspect 
DAML-OIL will achieve the kind of market dominance currently enjoyed 
by Apache in the webserver market.

Ciao,
Rex
-- 
Rex Brooks
Starbourne Communications Design
1361-A Addison, Berkeley, CA 94702 *510-849-2309
http://www.starbourne.com * rexb@starbourne.com