Re: [SKOS] A new proposal for ISSUE-39 ConceptualMappingLinks

Hi Alistair, all,

Sorry to take so long to reply.

Miles, AJ (Alistair) wrote:
>> My interpretation of the fact that there is development of a 
>> skos mapping vocabulary, which has been further confirmed by 
>> Antoine's email, is that the semantic relationships defined 
>> in the skos core [2] are to be used only for relationships 
>> between concepts in the same scheme.
>> However, this is not explicitly stated in the text of the skos core.
>> Will this be changed in the next version of the skos core?
>>     
>
> We don't know yet, it's under discussion. Do you have any preference?
>   
I very much believe that there is a difference between the relationships 
that take place within a vocabulary and those that take place between 
vocabularies.

In the first case (intra-vocabulary relationships), all the concepts 
form part of a coherent whole and the relationships can be seen as a 
statement of fact. The relationships form part of the vocabulary, 
meaning that if you make use of the vocabulary then you accept all the 
intra-vocabulary relationships.

In the second case (inter-vocabulary mappings), the mappings between the 
concepts are "fuzzier" and are more a statement of one person's, or 
group's, beliefs. The vocabularies can be used without accepting the 
statements made about the mappings.

We are working with the idea that several different groups will declare 
their mappings between vocabularies, and users can load in the one that 
they agree with. Over time, it may be the case that one or two sets of 
mappings become the de facto standard but that there will be an 
evolution process to this state.

>> A question I would like to raise is how can I specify a 
>> mapping between a collection in one vocabulary and a concept 
>> in another? It really is the collection as a whole that 
>> matches the concept. However, the collection becomes an 
>> anonymous node in the rdf. Is it the case that each member of 
>> the collection should be specified as a narrowMatch of the concept?
>>     
>
> It'd be great if you could give us some more detail on this particular requirement, e.g. the actual concepts you want to map between.
>   
Here is an example taken from mapping a vocabulary of journal keywords 
(AAKeys) [1] to the astronomical outreach imagery meta data vocabulary 
(AOIM) [2].
In the journal keywords there is the following:

    AAKeys:Stars
        ...
        NT AAKeys:[Binary stars]

where the collection AAKeys:[Binary stars] has the members

    * AAKeys:Binaries close
    * AAKeys:Binaries eclipsing
    * AAKeys:Binaries general
    * AAKeys:Binaries spectroscopic
    * AAKeys:Binaries symbiotic
    * AAKeys:Binaries visual

and each of these has no other relationships.

In the aoim vocabulary, there is the following:

    AOIM:Star
        ...
        NT AOIM:[Grouping]

where the collection AOIM:[Grouping] has the members

    * AOIM:Binary
    * AOIM:Triple
    * AOIM:Multiple
    * AOIM:Cluster
          o AOIM:Open
          o AOIM:Globular

The concepts AOIM:Open and AOIM:Globular are narrower terms of AOIM:Cluster.

The collection AAKeys:[Binary stars] is an exactMatch for the concept 
AOIM:Binary.

There are about half a dozen similar cases in mapping these two 
vocabularies.

Hope you are able to understand this.

Alasdair


[1] 
http://www.aanda.org/index2.php?option=com_content&task=view&id=170&Itemid=191
[2] http://www.ivoa.net/Documents/latest/AOIMetadata.html

-- 
Dr Alasdair J G Gray
http://www.dcs.gla.ac.uk/~agray/

Explicator project 
http://explicator.dcs.gla.ac.uk/

Office: F161
Tel: 	+44 141 330 6292

Postal: Computing Science,
	17 Lilybank Gardens,
	University of Glasgow,
	Glasgow,
	G12 8QQ, UK.