Jack, Joanne, Leo, (01)
I just wanted to make some comments on some of the points: (02)
JR
> The first thing they want to hear is how they can gain better
> knowledge exchange and choice making throughout their extended
> enterprise.
>
> The second thing they want to hear is how they can ascend
> from coordination to cooperation to collaboration to co-learning
> to co-evolving. (03)
There is no question that we have to "Sell the problem before
we can sell the solution," as the IBM salesmen were taught. (04)
And we're not going to get very far by getting on soap boxes
about our pet technologies. We have to start where the
customers are. Unfortunately, there is no such thing as
an "average" or "typical" customer. So a one-size-fits-all
sales pitch isn't going to get very far. (05)
JL
> We need to sell on the Flexible Data Model... (06)
I certainly agree on flexibility. But I looked at Segaran's
slides, and there was nothing new -- except a huge amount
of pizzazz. (07)
I'll admit that he is an enthusiastic salesman. But I got
turned off when I looked at his blog and was hit by five
occurrences of 'awesome' in half a page. (08)
JL
> that is underlying the ontology (e.g. RDF) (09)
RDF is not an ontology. It is just one among many, many
different data models. (010)
Ontologies, by nature, are independent of any or every
data model on the planet -- they are (or should be) the
apotheosis of flexibility. (011)
JL
> with the flexible data model, new "fields" can be added
> without breaking the schema... (012)
Yes, of course. But that is why the *conceptual schema*
(ontology) was decoupled from the *data schema* back in 1978.
Unfortunately, certain database vendors, who had a vested
interest in keeping their customers tightly coupled, fought
very hard to keep flexibility out of the standards. (013)
Leo
> We'll always need both terminologies (ways to refer, i.e.,
> the terms, words/phrases, labels) + ontologies (the concepts,
> referents/categories of referents, i.e., the representation
> of the meaning of those terms). (014)
Yes, indeed. Just a few points: (015)
1. Terminologies come first, and for many applications they
are sufficient for data interchange. (016)
2. Ontologies are essential for precise reasoning at the detailed
level, but we need to be careful about not getting too detailed
too soon. (That was a lesson that Lenat learned about Cyc,
which led to their moving many axioms from the upper levels
into the microtheories.) (017)
3. Both terminologies and ontologies are completely independent
of the data model. You can use them with any kind of database,
relational, network, or whatever. (018)
Leo
> Most of the "metadata" wars that we have all experienced are due to
> the conflation of these two notions, term and concept, label and meaning. (019)
I partly agree. But the main problem is that most of the people who get
hung up on those issues have no understanding of elementary logic. (020)
JR
> Next comes the General Purpose Set Theoretic Processor which when
> implemented in RAM will give constant throughput at a Gb/sec regardless
> of the size and complexity of the ontology. (021)
I agree that new technology can make a huge difference. (022)
John (023)
_________________________________________________________________
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/OntologySummit2011/
Community Wiki: http://ontolog.cim3.net/cgi-bin/wiki.pl?OntologySummit2011
Community Portal: http://ontolog.cim3.net/wiki/ (024)
|