W3C home > Mailing lists > Public > public-lod@w3.org > March 2013

Linking to non-RDF datasets

From: Alasdair J G Gray <Alasdair.Gray@manchester.ac.uk>
Date: Tue, 12 Mar 2013 10:18:07 +0000
Message-Id: <27FF63E1-7D56-4F8B-AD0A-AA146710E18B@manchester.ac.uk>
To: Linking Open Data <public-lod@w3.org>, semantic-web@w3.org
Hi All,

We are making extensive use of the VoID vocabulary [1] in the Open PHACTS project [2] to describe our datasets.  We are currently deciding how to model a recurring use case of needing to describe non-RDF datasets and manage linksets to them.

In the VoID vocabulary, a dataset is defined to be [3]
A set of RDF triples that are published, maintained or aggregated by a single provider.

Since all predicates are defined with a domain/range of void:Dataset, this would mean that it would be incorrect to use them for any dataset that is not a set of RDF triples. However, this usage is becoming common.

Should we go ahead and use the predicates despite this inaccurate interpretation of the non-RDF dataset? 

Is there another vocabulary that allows for the modelling of linksets that does not restrict the dataset to a set of RDF triples? I am aware of DCAT [4] but do not see suitable linking predicates.

Should we develop a set of super-properties that do not have the domain/range restrictions?



[1] http://www.w3.org/TR/void/
[2] http://www.openphacts.org/
[3] http://vocab.deri.ie/void#Dataset
[4] http://www.w3.org/TR/vocab-dcat/

Dr Alasdair J G Gray
Research Associate
+44 161 275 0145


Please consider the environment before printing this email.

Received on Tuesday, 12 March 2013 13:20:25 UTC

This archive was generated by hypermail 2.4.0 : Thursday, 24 March 2022 20:30:03 UTC