ExpertFinder 2fEFW2007 2fSlot3 5fMappingBetweenRDF
From FOAF
Responsible to report: MaxVoelkel
This group discussed the need (or no need) for mappings between overlapping RDF vocabularies. Something as powerful as XSLT doesn't exist for RDF, SPARQL construct queries might be a starting point, but are not sufficiently expressive, XSLT could be used on canonical RDF serializations as another approach, or alternatively any adhoc scripting approach, hidden behind a service. materialization vs dynamic execution of mappings was discussed as another topic: The disadvantage od scripting approaches is their non-declarativity, so mappings which are only available as a program are only executable, but could not be adaptively rewritten/combined e.g. for query rewriting.
Note that there are controversial arguments for and against using SPARQL as a rules language, see:
* http://danbri.org/words/?p=63 * http://www.polleres.net/publications/GIA-TR-2006-11-28.pdf * http://www.uni-koblenz.de/~aggrimm/arbeitsberichte/arbeitsberichte_3_2007.pdf
ACTION: Max, Maria Carmen, Emanuele, Uldis and Axel agreed to try to carve out the problems and existing approaches a bit better, possibly leading to a joint paper. The
[:ExpertFinder/mappings:mappings] page could serve as a starting point to collect this information.
