On Tue, 2011-08-23 at 17:22 -0700, conor dowling wrote: [ . . . ] > One last thing, slightly off the thread but still on topic I think. I > don't see any reason to mix up "human readable" and "machine > processable". One possibility for a patient model update, one that > bypasses the need for buy-in by everyone, irrespective of use case, is > to call out the need for a model of description purely for machine > processing, one without the "we'll XSLT the patient record in the > doctor's browser". While the current standards lead to human-readable > data-dumps, a stricter parallel track could take the best of current > standards and re-state them in OWL to deliver machine-processable > health data exchange, > Agreed. One of the nice things about semweb technology is that presentation ontologies can peacefully co-exist with other ontologies that are intended more for internal processing. > >
-- David Booth, Ph.D. http://dbooth.org/ Opinions expressed herein are those of the author and do not necessarily reflect those of his employer.