W3C home > Mailing lists > Public > public-xg-prov@w3.org > June 2010

Summary of last week's discussion re the vocabulary mapping

From: Olaf Hartig <hartig@informatik.hu-berlin.de>
Date: Thu, 3 Jun 2010 11:51:26 +0200
To: public-xg-prov@w3.org
Message-Id: <201006031151.27850.hartig@informatik.hu-berlin.de>
Hey,

I cannot make this week's call (still at ESWC). However, as a starting point 
to continue the discussion on the provenance vocabulary mappings at this 
week's call, here's a summary of the discussion we had of last week.

The first (and nearly completed) task was to define which terms of each 
vocabulary are related to a common set of terms and what this relationship is. 
We represent this information in the mapping table on this wiki page:

 http://www.w3.org/2005/Incubator/prov/wiki/Provenance_Vocabulary_Mappings

For the common set of terms we selected the provenance-related terms from OPM 
and to represent the relationships we choosed SKOS. Furthermore, we list (and 
that still has to be done for some vocabularies) the provenance-related terms 
of each vocabulary that are not mentioned in the mapping table. This listing 
will be a useful basis for a gap analysis.

Based on this work, we (as the group members who worked on this mapping) 
raised the following two questions that we wanted feedback from the whole 
group:

1) Should the information in the mapping table be represented as an RDF graph. 
Such an RDF graph could be used for some automatic alignment of provenance 
descriptions that are based on different vocabularies.

During last week's call there was a general agreement that such an RDF graph 
would be useful. However, Ivan suggested to be very careful and discuss each 
of the statements that we would put in this graph with the original authors of 
the corresponding vocabularies. That might turn out quite some challenge. I 
proposed to add at least these statements that correspond to vocabularies 
developed by people who are part of our group.

2) How should we proceed? Should we consider the exercise of defining a mapping 
as done? Or should we extend the set of common terms in the mapping table 
based on an analysis of the terms not mentioned in the table? Would such an 
extension still be in the scope of our work as an incubator group?

During the call last week there was no consensus on these questions. Yolanda 
proposed to extend this work in a direction were we additionally align the 
terms (only the common ones are all provenance-related terms from all 
considered vocabs?) with our requirements.

That's what I remember from last week's discussion. Please add, if I forgot 
something.

Greetings from Crete,
Olaf
Received on Thursday, 3 June 2010 09:52:36 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 3 June 2010 09:52:36 GMT