Lists Home |
Date Index |
The problem is it isn't contract, but contracts.
RFP by RFP. It is great if they can all reference
one ontology, but for that to work, that ontology
has to be the sum of their requirements; whaddaygit?
Another bloated specification. Just whining, here.
It isn't that the ontology drifts: it is that
meaning drifts. Will I accept a noise ratio
of 5 to 1? Sure. Sobriety rules. One can't
count on a large non-local community being sober
all the time in all of the places where they
make their decisions. So not just sober choice,
but well-considered application. That is as good
as it gets and why many said that frictionless
computing was/is nonsense, so YMMV.
Don't get me wrong. We're very happy to get
standards for the codelists we use. Stuff them
into an enumeration and let us suck them via an
XMLReader right into the database, then to the
dropdown. Very happy indeed. But the real trick
is to in near real time detect that a user in a
particular context chose the wrong value from that
list. This is when the semantic stuff starts to
have more value.
From: Jeff Rafter [mailto:email@example.com]
> Do the best you can but no one
> can make time or meaning stand still. YMMV.
Sure they can, in the form of contracts. Essentially that is what OWL is
for right-- a contract about the nature/meaning of a particular piece of
information? Sure, those considerations will change over time but that
is what versioning is for?
Semantic drift is to be expected, and I'll grant that it is a problem
but that doesn't mean it makes the whole process useless. I know that
the fidelity of an MP3 recorded from a CD and an old cassette are two
wildly different things. I know that converting the MP3 to another
format and back will likely involve some loss-- but it doesn't mean that
the information is useless, I just have to approach soberly.
Code lists are great, shared code lists are more great-- but for each
level you go out you have to keep in mind that there will be some
lossiness. Fine. Still, sign me up-- if I have a program that can auto
map 1800 out of 2000 fields reliably, I'll use it.