On 4/4/06, Azamat <abdoul@xxxxxxxxxxxxxx> wrote: (01)
> > I'm sorry, but the majority of web applications I see today have
> > little or no grounding in traditional ontologies, upper or otherwise.
> > I do believe that the descriptive capability that the SW languages
> > introduce offers the potential for more useful web applications.
>
> This capacity comes from a core set predicable relations of: ''definition'',
> ''class'', ''individual'', ''property'', ''sameness'', ''difference'', and
> ''inheritance''. It is nothing to do with this sort of formal language. (02)
I don't disagree, just wished to make the point that a formal language
(of any level of sophistication) isn't a prerequisite for useful web
apps. (03)
> > This seems reasonable. If I wish to take my heavy vehicle over a
> > bridge, I would need firstly to discover the maximum weight the bridge
> > could support - or to put it another way, that property shared by all
> > objects in the set of things that could get over the bridge. Next I
> > would need to consider the weight of my individual vehicle.
>
> It is not so complicated. In UFO, there are basic ontological assumptions
> for making a situational reasoning, such as
> There are substances (objects) in the world;
> The world can be in different states;
> Objects have states;
> Object are subject to changes;
> Changes cause other changes as effects;
> A state is the union of all the values of the properties which an object
> (system) has at a certain time, etc.
>
> Basing on such axioms, your reasoner may conclude that your car is a
> physical system ( a vehicle) marked by specific physical qualities and
> quantities, as mass and weight, position, velocity, and all the properties
> take some range of values, with upper limit and lower limit. The same
> reasoning applies for the bridge. (04)
Such a system would be nice to have, but you can't deny this would
involve more effort than the simplistic model I described. All things
being equal, the "natural selection" of the web environment tends to
favour low-effort solutions over more complex ones, where there isn't
any (perceived) difference in their utility. (05)
Semantic Web technologies are a hard sell to people that have achieved
results, solved problems, using approaches without any real formal
base whatsoever. That is unless they provide visible benefits.
Fortunately such benefits are demonstrable, and I think we're going to
see increasing numbers of systems using these technologies - the
deployment phase has started. (06)
I'm personally unconvinced of the advantages of a UFO kind of approach
in itself, but that aside, for such a system to gain widespread
acceptance/adoption it must first be built and demonstrate benefits. I
would suggest that the easiest way of creating such a system would be
to build on SemWeb tech. (07)
> > For the reasons given at the start, I'm not really qualified to talk
> > in terms of the nature and fundamental kinds of things. But I am
> > qualified (by experience) to say it is possible to usefully represent
> > and reason about reality with the kinds of languages found in the
> > SemWeb stack.
>
> See above. Here is a main problem: it's a push-down stack without any
> ontological groundwork.
>
> > For example, software can be used to determine whether two people are
> > likely to have an opportunity to meet each other at a conference using
> > very simply constructed representations of the information, like the
> > class of Person and Location and properties like date/time. (see [1]).
> > There's no need to anchor this information in the fundamental nature
> > of reality,
>
> Actually, there is; for your real-life applications should 'understand' what
> does it mean to be a person, to have a location, etc. (08)
Not so. All it needs is a model that holds for the pertinent
information. To determine whether two people can meet there's no real
need to explicitly model a person as anything other than a thing that
can be at a location; a location as either a named place or geo
coordinates; time as a set of conditions for which coincidence can
hold. (09)
> > However, through Semantic Web languages, information expressed using
> > these terms is potentially compatible with that expressed in any other
> > formalisation of the world.
>
> You may have your individual ontological commitments specific a certain
> domain (your home, a domain-specific ontology). (010)
True. (011)
Important, these commitments
> are not casual but must comes form one underlying reality, unifying
> ontological context. (012)
I'd argue that the reality you describe, in the context of
machine-interpretable ontologies is but another (specific) domain. (013)
For then the application can tell your most meaningful
> information about you pets and wife, their properties, possible behavior,
> and relationships. That they are living beings, have life, can eat, have
> different moods, can love you, etc. (014)
As long as I take to notice when they're hungry, the other aspects
tend to take care of themselves ;-) (015)
Cheers,
Danny. (016)
-- (017)
http://dannyayers.com (018)
_________________________________________________________________
Message Archives: http://colab.cim3.net/forum/ontac-forum/
To Post: mailto:ontac-forum@xxxxxxxxxxxxxx
Subscribe/Unsubscribe/Config:
http://colab.cim3.net/mailman/listinfo/ontac-forum/
Shared Files: http://colab.cim3.net/file/work/SICoP/ontac/
Community Wiki:
http://colab.cim3.net/cgi-bin/wiki.pl?SICoP/OntologyTaxonomyCoordinatingWG (019)
|