[
Lists Home |
Date Index |
Thread Index
]
Jonathan Borden wrote:
> The problem with this analysis is that the world isn't _only_ about
> processes and processing -- there exists a declarative approach which has a
> fundamental fit with XML -- which fundamentally is, or describes, a data
> format (not a process description).
>
> The ontologic approach is such a declarative approach -- for example,
> nowhere in the OWL specifications is *any* writing about a "processing
> model". You may consider such ontologic descriptions as specious
> abstractions, but the whole point of this excercize is that much can be done
> *without specifying any processing*.
Strong agreement. But let's remember that you cannot have a theory
of content without a process model - something some luminaries
involved now in the semantic web/ont efforts realized a long time ago.
>
> We have only data and descriptions of data. When I was first discovering
> XML, I recall reading the statement that XML is appropriate when the useful
> life of the data is expected to be longer than the useful life of the
> software that processes it. It seems to me that descriptions of data can
> similarly outlive the processes that derive from these descriptions.
So much of what we want to know is abductive, and cannot be
discovered by writing everything down and running theorem provers
over the data. I think Walter is trying to remind us that processes
and process models must play a big part - on the other hand there's
no question that at the moment there are far too many conditional
blocks in systems code ('business logic' in the jargon) that would
be better off placed in declarative sentences for comsumption from
pattern matchers to reasoners.
Bill de hÓra
--
"A year spent in artificial intelligence is enough
to make one believe in God." -Alan Perlis
|
