This switch in threads is satisfying. In my thesis proposal [1], I
defined federation, the process, along the lines of its dictionary
definition, to bring together, but to do so without filters (perhaps
other than those associated with legal issues). Let me contribute a
few thoughts. (01)
A knowledge federation, as I see it, is a collection of points of
view, justifications for those points, arguments, and much more. (02)
It is always useful that any point of view be anchored in concepts
defined in some ontology, but not all artifacts in the federation will
be so grounded. (03)
My own work is to build topic maps, not in the traditional XML DTD my
book on topic maps talks about, but instead, in the way prescribed by
the topic maps reference model (TMRM) which simply expects you to
capture topics in proxy objects which are containers for what GOFAI
calls slots and fillers. I read ontologies of various kinds into one
space, then map them to topics in the topic map to provide anchors
where needed. I use the Bigdata triplestore, but not OWL, for my
topic maps. My most recent work explores the intersection of
conceptual graphs and topic maps. (04)
All that to federate our world views. The rubric for that is
"knowledge gardening" [2]. (05)
Jack
[1] http://kmi.open.ac.uk/publications/techreport/kmi-10-01
[2] http://slideshare.net/jackpark/ (06)
On Mon, Jan 23, 2012 at 10:25 AM, Bobbin Teegarden <teegs@xxxxxxxxxxxxx> wrote:
> Cory’s point – that we live in and need to design for a federated world –
> where domains end up being more a point of view than a definition, is key.
> I would also add that the federation is simultaneous, and could be massively
> parallel, such that events no longer can be viewed as singular from an
> architectural pov. With ontologies, we can wall the gardens while
> federating the surrounding forrest.
>
>
>
> But, as mentioned earlier, there are some – only some – things you can say
> with OWL. And while we have reasoners, they may check for consistency etc,
> but they go off to lunch if you use inverseOf.
>
>
>
> And in designing a system (or systems of systems), chasing a live, morphing
> system around with no versioning and imprecise time specification is still a
> challenge. Process is quite missing; ontologies still prefer the static.
>
>
>
> The biggest challenge, though, may be the lack of better tools, especially
> visualization of the multi-dimensional worlds we can now capture.
>
>
>
> I’m not sure if these comments are in one Engineering Track or several – but
> I’ll add them to the list of considerations.
>
>
<snip> (07)
_________________________________________________________________
Msg Archives: http://ontolog.cim3.net/forum/ontology-summit/
Subscribe/Config: http://ontolog.cim3.net/mailman/listinfo/ontology-summit/
Unsubscribe: mailto:ontology-summit-leave@xxxxxxxxxxxxxxxx
Community Files: http://ontolog.cim3.net/file/work/OntologySummit2012/
Community Wiki: http://ontolog.cim3.net/cgi-bin/wiki.pl?OntologySummit2012
Community Portal: http://ontolog.cim3.net/wiki/ (08)
|