- From: Michael Brunnbauer <brunni@netestate.de>
- Date: Wed, 8 Feb 2012 18:31:13 +0100
- To: semantic-web@w3.org
hi all On Fri, Feb 03, 2012 at 02:35:57PM -0500, Tim Finin wrote: > * A General Framework for Representing, Reasoning and Querying > with Annotated Semantic Web Data, Antoine Zimmermann, Axel > Polleres, Nuno Lopes and Umberto Straccia http://www.websemanticsjournal.org/index.php/ps/article/view/216 I found this paper quite fascinating: The standard RDF semantics is replaced by a many valued logic with truth values that build an algebraic semiring. The truth values represent information about the minimum extent the information given in the triple is considered to be true (e.G. temporal, provenance or trust extent). The values do not have to be simple numbers but can be any data type that builds a semiring - for example sets of time intervals or propositional formulas built from atomic propositions about provenance. The authors show how combinations of annotation domains like temporal+fuzzytrust can be handled automatically. SPARQL is extented to support the annotation mechanism and quad stores are suggested as storage where the context part of the quad is a typed literal representing the annotation or truth value. Apart from the fact that this breaks most of the semantic web standards and tools (how many quad stores do not support typed literals as context ?), I like this. It would be interesting to know why the authors only use a subset of RDFS and if there are obstacles to extending this to OWL full. I was thinking about how to achieve the same with named graphs but this seems to get awkward with reasoning: Every named graph can only have one annotation/truth value that is valid for all it's triples. If you use a reasoning rule to deduce a triple out of n other triples from several graphs, the annotation/truth value of the deduced triple is computed from the value of the n other triples and potentially new. As deduction rules are used recursively, you end up with a plethora of annotation/truth values and each one requires a separate named graph with deduced triples in it. Are these dimensions (temporal, provenance or trust, etc.) so important that we have to implement them as such (and reinvent almost everything) or can we live with other solutions that probably will be crippled in some way ? Regards, Michael Brunnbauer -- ++ Michael Brunnbauer ++ netEstate GmbH ++ Geisenhausener Straße 11a ++ 81379 München ++ Tel +49 89 32 19 77 80 ++ Fax +49 89 32 19 77 89 ++ E-Mail brunni@netestate.de ++ http://www.netestate.de/ ++ ++ Sitz: München, HRB Nr.142452 (Handelsregister B München) ++ USt-IdNr. DE221033342 ++ Geschäftsführer: Michael Brunnbauer, Franz Brunnbauer ++ Prokurist: Dipl. Kfm. (Univ.) Markus Hendel
Received on Wednesday, 8 February 2012 17:34:31 UTC