- From: Mark van Assem <mark@cs.vu.nl>
- Date: Thu, 2 Dec 2010 09:53:53 +0100
- To: Emmanuelle Bermes <emmanuelle.bermes@bnf.fr>
- CC: Karen Coyle <kcoyle@kcoyle.net>, public-xg-lld <public-xg-lld@w3.org>
Hi Emma, In our end report I assumed we need to have consistent language in the UCs contained therein. That's why we started identifying which things in the Goals section of the UCs that were being said over and over but in different wording. In that sense what we did mirrors what the UC owners have written, not our own analysis of what the goals really are. Thus what we found may be goals, requirements, howtos, ... Basically I think we are still wrestling with a larger question: what are we going to do with 40+ UC texts in our end document? Creating scenarios and goals is kind of an abstraction approach. An option is to use these to provide a really condensed description in the end document, and put the full UCs in an appendix. > For instance : > - DESCRIBE: describe a resource and/or provide context for a resource. > This one is probably related with FIND/SEARCH and DISCOVER/SUGGEST but > I don't know how so. That depends on what you mean with describing a resource and providing context for a resource... Does it mean that you add extra information, ie extra triples, which are not in the original dataset? > - A goal we've identified would be "facilitate cross-domain modelling > & mappings" : I'm not sure if this > one would correspond to the current "MAP" : "create equivalence > relationships (owl:sameAs, skos:closeMatch) between value vocabularies > or data items." Maybe we should split between MAP-VALUE for value > vocabularies, and MAP-METADATA for metadata elements sets or metadata > vocabularies (we still need our fixed terminology here !) "facilitate" would consist of providing interface support to add mapping statements? Your suggestion for MAP-VALUE and MAP-METADATA can also be covered with our suggested qualifier approach; eg MAP (value), MAP (metada). The reason for qualifiers is that every extra option/attribute of a goal would end up getting us a too large set of goals. I don't know what's meant with "cross-domain modelling"... Things may become clearer if you publish your goals in relation to your UCs goal texts. Best, Mark. > On Tue, Nov 30, 2010 at 12:11 PM, Mark van Assem<mam500@vu.nl> wrote: >> Hi Emma, Karen, >> >> In the last telecon you indicated that the current descriptions of UC Goals >> [1] were not clear enough for you to apply them in your cluster's UCs. >> >> Is there any way we from the DO cluster team (Jodi, Asaf and me) who came up >> with them could help to clarify? Which ones are problematic? >> >> Best, >> Mark >> [1]http://www.w3.org/2005/Incubator/lld/wiki/Goals >> > > >
Received on Thursday, 2 December 2010 08:54:21 UTC