Why RDF Tech/ Why NOT UCD-ish approach (Was: Re: Vocabulary: Ontology
thomas at astro.umd.edu
Tue Sep 11 14:38:10 PDT 2007
Everyone else is batting about platitudes to make their point, I
might as well do the same..
On Tuesday 11 September 2007 5:06:48 pm Rob Seaman wrote:
> Rather, shouldn't the VO be seeking creative technologies whether old
> or new? An argument for RDF surely should begin by emphasizing how
> sexy an option it is.
Sure "sex sells", but its not the basis for a good long term relationship,
and you often wake up in the morning regretting...
But more seriously..
Look, I believe we all want to answer the question: "Where are the
scientific data of interest in the VO". Frankly, I don't see how we can
achieve that goal without doing one or the other of the following:
1. Implement some semantic technology which allows the machine to
handle a good bit of the figuring out where the data are (e.g. inferences)
2. Forcing a single standard to apply across the VO so that all data are
The second option has been tried with the UCDs, and while there have been
some success, its limitations are fairly glaring as well. It wont answer the
question "Where are my data of interest" at all. If it did, then why are we here?
Why aren't we simply adding more UCD's? No, we have been down the UCD
route, and its time to choose a different technology.
Other than the present approach is not adequate, how much more justification
do we need for using a tried and true, well supported, more advanced than we
presently have, which does the things we need, technology than that (hint hint:
More information about the semantics