As soon as we agreed on something like "score" being a standard, why not
make it an optional <srw:score> tag directly below <srw:record>. I had
the impression that extraRecordData was meant for things that are not or
not yet agreed upon in the standard or am I wrong?
>>> [log in to unmask] 11-6-04 8:24:55 >>>
Ray Denenberg, Library of Congress wrote:
>>>I do not have SRW on my fingertips, but I belive that there is no
>>>'reserved' tag inside the 'extraRecordData' tag, and if it is not
>>>defined by the standard, we are back to the problems of approach
>I think we have all the tools we need to do this (and we don't need
>"reserved" tags), but we really should nail it down a little better
>this is important for the metasearch initiative.
> I just took a look and Rob has already defined "score" in his
>http://srw.cheshire3.org/schemas/rec/1.0/, schema at
> <rec:score rec="info:srw/schema/2/rec-1.0">.73</rec:score>
>I suggest that we agree on a normalization range, say 0-1 as
>modify the definition as such.
This I do really appreciate.
> (I don't think there's a need to be able to specify different
>normalizations, and that just makes the metadata more complex -- e.g.
><extraRecordData> <rec:score rec="info:srw/schema/2/rec-1.0">
> <normalization> 0,1</normalization>
></rec:score> </extraRecordData> )
This will definitely be overkill. If such diversity should be allowed,
it should only be allowed on a per-server basis, that is , this
information should go into
a <extraServerData> tag once, not in every <extraRecordData>.
>We still need a way to request the metadata.
>For example, we could define:
> <rec:score xmlns:rec="info:srw/extension/1/record-metadata"/>
>where the extension identifier explicitly means you're requesting the
>specified metadata, in this case score.
I am too green in SRW to have an opinion on this.
Cheers, Marc Cromme, Index Data
>the above though, only if we establish an association between this
>extension id and Rob's metadata set.
>More flexibile would be:
>where info:srw/schema/xx/yyyyy identifies some metadata namespace
>score is an element of that namespace.
>Is this what we have in mind?