On tis, 2007-02-06 at 14:17 +1030, Jon Mason wrote:
> 6. Each of the 3 models are far more clearly presented than in the
> previous version -- in particular, the roles of "identification",
> "representation", & "description". This also raises another issue:
> many of the more simplistic & commonplace definitions of "metadata"
> define it as "data that describes other data". Would a more precise
> "simplistic" definition then be "data that is described, identified,
> or represented by other data"?
Thanks for the positive feedback on the improvement to the models that
is a central point, and very good to hear.
I think you wording of a metadata definition gets the referencing
backwards... I think maybe
"data that describes, identifies, or represents other data"
is more correct, even though I'd be careful to generalize this to:
"data that describes, identifies, or represents resources"
I must say I'm not sure "describe, identify, represent" really captures
all "aboutness" in metadata, but it's worth a thought...
Also, I'd hasten to modify that to the following:
"Machine-processable data that describes, identifies, or represents
resources"
I guess I could live with that.....
/Mikael
--
<[log in to unmask]>
Plus ça change, plus c'est la même chose
|