The Resource Description Access (RDA) Linked Data Forum took place at the 2018 ALA Annual Conference in New Orleans, Louisiana, at the Morial Convention Center on Monday, June 25. James Hennelly (director of ALA Digital Reference) introduced the session’s speaker and moderated the question and answer session following the presentation.
Gordon Dunsire (chair of the RDA Steering Committee) delivered a presentation titled “RDA Linked Data and the New RDA Toolkit.” Dunsire began his presentation by emphasizing that the Beta version of the Toolkit released on June 13, 2018, is not complete or usable in its current form. The expected completion date has been extended to December 2018 for the English version of the Toolkit. Language translations and completion of accompanying Policy Statements are expected to extend into 2019. The impact of the Library Reference Model (LRM) and the need for a more responsive design and structure precipitated the RDA Toolkit Restructure and Redesign Project, or the 3R Project. The redesign allows the RDA Registry to supply content in response to community needs. Indeed, making the Toolkit more responsive is an integral part of RDA’s strategy for international, cultural heritage and linked data communities.
The development of entities, elements and terminologies within the RDA Registry is almost complete. New LRM entities include Agent, Collective Agent, Nomen, Place and Timespan. Many of the attributes in the previous versions of RDA are now relationships. The new Toolkit will also include more inverse relationships than the previous Toolkit. In short, the LRM is optimized for linked data with 13 entities and over 1,700+ elements. The RDA vocabularies are published via GitHub, and the RDA Registry includes the latest version of the RDF vocabularies. The RDA Registry is a user-friendly way to see RDA as linked data.
A new feature of the RDA Toolkit includes consistency in recording instructions with four methods of recording data values which include:
- unstructured description
- structured description
Unstructured description includes the transcription of free-formed notes, such as the transcription of a name or title. An unstructured description is a string with no range specified. Structured description uses multiple values with a string encoding scheme, such as an access point. A structured description is also a string but with a range indicated by the accompanying encoding scheme. An identifier is a machine-readable local appellation that is unique within the context of the schema to which it pertains, such as an ISSN. An identifier is a literal with a specified range. An IRI is also a machine-readable appellation that is intended to be global and truly unique. An IRI is a thing, not a string.
Dunsire concluded his presentation with a summary of RDA releases and semantic versioning. The version number of the last release of the RDA Toolkit in April 2017 was 2.7.3. Dunsire advised that the July 2018 RDA Toolkit release will be version 3.1.0. The first number in the version changing from a “3” to a “2” indicates that the semantics between the April 2017 and July 2018 versions of the toolkit will be broken. The reason for this break is the change and narrowing definition of the Person entity, which now excludes spirits, fictitious persons, etc.
In short, the new RDA toolkit is a package of data elements, guidelines and instructions for creating metadata for user-focused linked data applications designed to meet the resource description and access needs of the twenty-first century. For more information, select RDA Steering Committee presentations are available on their website.