[om-list] Subject / Schema Oriented Computing

[Thomas L. Packer]

Mark

	Sounds related to issues in agent-based AI, or at least in agent 
modelling of natural language comprehension, generation, and 
especially dialogue.

tomp


On 2003 October 07 Tuesday 13:41, Mark Butler wrote:
> Subject / Schema Oriented Computing
>
> I am working on subject oriented computing these days.  What
> is that you say?  Well, the foundation of object oriented
> modelling is the idea that you need only deal with a single,
> accurate representation of objective reality.  The language of
> object oriented modelling is not suited to situations where
> you have to keep track of what several different parties
> beleive to be true about the same object, let alone what one
> party believes another party believes to be true.
>
> Given that object oriented computing is suited for an
> objective perspective where belief is irrelevant, I thought
> subject (or schema) oriented computing was a good name for any
> modelling technique that revolves around tracking multiple
> perspectives, especially perspectives on perspectives.
>
> For example, if you are a negotiator, you need to know three
> things - what you know to be true, what your client believes
> to be true, and what the other side believes to be true. 
> Traditional object oriented modelling (not to mention
> scientific reductionism in general) only deals with the first.
>
> These are the working definitions that I came up with today,
> for a model of genealogical knowledge I am developing:
>
>  Schema
>
>    A qualified interpretive model of knowledge and belief.
> What an author personally believes or what he beleives another
> to beleive.
>
>      Root Schema
>
>        Direct beliefs of a primary author
>        (e.g. what I believe now)
>
>      Child Schema
>
>        Beliefs or assertions held in the past or regarding the
> beliefs of others.
>
>          Current Foreign Schema
>
>            A child schema containing beliefs about the current
> beliefs of another
>            (e.g. what I believe another believes). Also known
> as a diplomatic schema.
>
>          Historical Foreign Schema
>
>            A child schema containing beliefs about the beliefs
> of an other at some time
>            in the past or as represented in some creative work
>
>            (e.g. what I believe another beleived at some time
> in the past)
>
>            - foreign schemas may in principle be repeated to
> arbitrary depth (he said that she said that he said...)
>
>          Historical Personal Schema
>
>            A child schema containing personal beliefs held at
> some time in the past or as represented in some personal
> creative work (e.g. what I believed when I was young)
>
>
> Schema Properties
>
>       The author of a schema is the belief modeller,
>       The author of a child schema is generally the same
>       as (or closely related to) the author of its parent
> schema.
>
>       The source of a schema is the subject who holds the
> beliefs being modelled. The source of a child schema may have
> no relation to the source of a parent schema.
>
>
> Subject or schema oriented computing is major overkill for the
> natural sciences, of course, but it appears to be very useful
> in the social sciences and in the humanities, at least for the
> realists among them, anyways.
>
> Comments anyone?
>
>    - Mark

-- 

ciao,
tomp at BYU
--------------------