Patrick -On Feb 17, 2013, at 10:17 AM, Patrick Cassidy wrote: **But**, if you are
designing a top-level ontology that can integrate other **independently
developed** top-level ontologies
I would argue that for all intents & purposes, INTEGRATION is not possible.
The realistic goal should be INTEROPERABILITY rather than integration.
The silos are in place & they will not move. The only cost effective way forward is to learn how to get the silos to effectively talk to each other.
As explained to me by a highly skilled USNA, USMC, data modeler...
If you have a blank sheet of paper, have budget control & you're in COMMAND (not the same thing as control), then you can potentially have an integrated <something>.
In the real world, there is no such thing as a blank sheet of paper.. the greenfield days of [information] systems are in the distant past.
Just think... when someone walks into an Apple store to buy an iPhone & walks out in 15 minutes... do they feel the ground shake as those 40-50+ year-old AT&T/Verizon mainframe systems provision that 4oz handheld device?
Plus... I have noticed that when an outsider attempts to force unfamiliar language onto people who are already familiar with local jargon, it is extremely unlikely that the outsider/newcomer will win the day.
Classic example: KICKS vs CICS (that's an IBMer inside joke, likely to be totally opaque to non-IBMers)
|