Re: "Connect library data standards to the linked data space"

Hi,

I think I agree with Jodi: the technical notion of alignment (i.e., creating explicit links as statements) reads too limited here. I felt the looser meaning of "bringing into line" was more appropriate as it could cover technical alignment between different vocabularies, but also direct re-use of existing ones.

By the way I've removed "standards" from the heading (now "Align library data with the Linked Data space"). This paragraph should not be focused on standards, which are the highlight of the previous section ("For Library standards"). Maybe this can help finding a convenient new heading, as per your discussion below :-)

I'd follow on Jodi's proposal, and suggest: "Express library data by re-using or mapping to existing Linked Data vocabularies".

Antoine


>
> On 5 Sep 2011, at 18:11, Tom Baker wrote:
>
>> On Sun, Sep 04, 2011 at 03:10:55PM -0700, Karen Coyle wrote:
>>> I think "connect" is stronger than align. I think we want actual
>>> connections (links). - kc
>>
>> Of the two, it is the alignments that I think of as "actual links", and I do
>> not see the word "connections" (or "connect") commonly used specially to mean
>> "links" (or "link"). Therefore, between:
>>
>> "Connect library data standards to the Linked Data space"
>>
>> and
>>
>> "Align library data standards with the Linked Data space"
>>
>> ...the latter, to me, more strongly connotes "linking", while the
>> the former is vaguely evocative but less precise. Jodi indicated
>> agreement (though perhaps for different reasons).
>
> When I think about "aligning" I think about semantic alignment/crosswalking/etc.
>
> I think that we want to be using the *same* standards whenever possible.
>>
>> However, I'm sure Karen is not the only one who sees "connect" as stronger than
>> "align". I would not object to reverting to the former.
>>
>> What do others think?
>
> I'm rereading the paragraph here:
>
>
>       Align library data standards with the Linked Data space
>
> In order maximize linkability with other datasets, library datasets must be expressed using Linked Data terms -- properties, classes, and instances -- that have well-defined relationships to those used in the wider Linked Data space. This can be done in two ways: by using Linked Data vocabularies based on existing standards, such as ISO language names; and by defining explicit relationships ("alignments") between the Linked Data terms of the library world and those of other communities.
>
> "Use Linked Data standards to express library data" would be another way to express *part* of this. I think that's the part that connotes linking, as Karen notes.
>
> However, to my ear, this part means something more than links (which I'm thinking of as bare <a href="..."> links): "by defining explicit relationships ("alignments")"
>
> That makes me think:
> "Express library data standards with Linked Data vocabularies and by mapping to existing vocabularies"
>
> I still prefer "align" to "connect", but I'm flexible: feel free to revert if you prefer
>
> :) -Jodi
>
>>
>> Tom
>>
>>> Quoting Tom Baker <tbaker@tbaker.de <mailto:tbaker@tbaker.de>>:
>>>> I'm tempted to say "align" here, and I think that some vagueness in the
>>>> text can be resolved by spelling out that datasets are expressed using
>>>> properties, classes, and concepts [1]:
>>>>
>>>> Align library data standards with the Linked Data space
>>>>
>>>> In order maximize linkability with other datasets, library
>>>> datasets must be
>>>> expressed using properties, classes, and concepts that have well-defined
>>>> relationships to properties, classes, and concepts used in the
>>>> wider Linked
>>>> Data space. This can be done in two ways: by using Linked Data
>>>> vocabularies
>>>> based on existing standards, such as ISO language names; and by defining
>>>> explicit relationships ("alignments") between the Linked Data
>>>> terms of the
>>>> library world and those of other communities.
>>>>
>>>> -- Note: "Linked Data terms" in order to avoid a tedious
>>>> repetition of "properties,
>>>> classes, and concepts"...
>>>>
>>>> [1] http://www.w3.org/2005/Incubator/lld/wiki/index.php?title=Draft_recommendations_page_take2&diff=6133&oldid=6132 <http://www.w3.org/2005/Incubator/lld/wiki/index.php?title=Draft_recommendations_page_take2&diff=6133&oldid=6132>
>>
>

Received on Tuesday, 6 September 2011 10:37:21 UTC