You are here

Assessing Linked Data Repositories

The second speaker in the linked data panel at DHA 2012 is Steven Hayes, who begins by introducing the network model of representing relationships between entities. This model has been employed by the Heurist database system, which Steven says represents a new ‘linked data’ mindset in humanities research.

From the perspective of that mindset, how linked are our data? Steven presents a number of criteria for ‘proper’ linked data: are they available online, in machine-readable form, using non-proprietary formats, using RDF standards, and linked to other RDF repositories, for example (the linked data checklist proposed by Tim Berners-Lee)?

Steven now assesses a number of Australian data repositories against this checklist now – including, for example, the Dictionary of Sydney or the collection of Balinese paintings – and takes us through some of the conceptual difficulties in representing such diverse datasets.

Do such assessments against ideal checklists matter, however? If datasets remain less well-linked, are they still linked data? There is a need for further Web standard development, and a clear desire to provide more linked and linkable data – but exactly how this proceeds may not follow the predesigned course desired by ‘linked data’ advocates.