Frederic V. Hessman wrote:
>
> On 2008 Jan 31, at 21:28, Norman Gray wrote:
>
>> Presuming that we'll be wanting to add mappings between these
>> vocabularies (that being half the point of this project), this is
>> starting to look like a lot of work. Perhaps A&A, IVOAT and AOIM
>> would be an adequately large set to include in the standard after all.
>>
> No, I'd say we should include mappings between A&A and AIOM simply to
> show that it can be done. Leave the rest for the question of what
> vocabulary is going to be the best lingua franca. It's tempting to
> use IAU-93, but there are simply too many mistakes and things missing
> and we can't/shouldn't correct for this.
The big problem with the IAU-93 is that there are no definitions for the
terms. This is another reason we need to get the IVOAT, or at least an
early version of it, ready for use. However, there will ultimately be
the need for mappings between pairs of vocabularies as it may not always
be possible to express these by going through a central vocabulary.
As a point of interest, I am just about finished developing an application to aid the generation of mappings between a pair of vocabularies. Once I have ironed out a few glitches and got it conforming to the latest skos reference working draft I will issue it for use.
Cheers,
Alasdair
>
> Rick
Received on 2008-02-06Z12:28:54