Toolkit currently being reorganized for easier use. Also the RDA Registry and the Toolkit are in the process of building full synchronization for all languages.
RIMMF does an excellent job both mapping from and to MARC, virtually losslessly.
Resource Description and Access (RDA) has been around long enough that it sometimes seems old hat. In fact, it is still sometimes described in terms of 'cataloging rules' as if its early history as an outgrowth of AACR2 defines its scope and reach in the present. In fact, RDA has matured into a very modern, international metadata standard, including a rich and expressive structural vocabulary based on FRBR.
It's sometimes noted that catalogers still 'speak' with MARC numeric codes, reminding us how much the MARC Standard has informed our thinking. More than that is the continuing concern about our legacy data, and the recognition that MARC output will be required for many smaller and special libraries well into the future.
Much of our current experience with 'native' RDA as a cataloging format has been in the context of RIMMF (RDA in Many Metadata Formats) which has been used for the many 'Jane-athons' scheduled for the last 18 months. RIMMF easily and completely maps legacy MARC data into FRBR-based RDA, and can also output MARC and other formats after creation or editing within RIMMF.
RDA is rapidly becoming the standard of choice outside of the US, as the Toolkit instructions and the RDA Registry become available in a multitude of languages. In this program, Diane Hillmann, a member of the RDA Development Team, provides an update on RDA progress, with special emphasis on its ability to serve as a lossless transition from MARC.
This is a partial screen-shot of the OMR view of the RDA Media type “microform”.
The term label, definition, and scope note in multiple languages are linked to the URI for the term. This basic feature of RDF, the separation of strings (labels, etc.) from things (URIs), is fully exploited by the RDA Registry.
Value vocabularies are represented using SKOS (Simple Knowledge Organization System) elements. SKOS allows one preferred label per language, and as many definitions and notes as are required. RDA uses SKOS elements for broader, narrower, and term or concept equivalence to represent semantic hierarchies and maps.
The RDA Toolkit is currently available in French, German, and Spanish, with Finnish and Italian translations scheduled for the next series of updates. A separate translation is published in China. The RDA Translations Policy requires all full translations of the Toolkit to be represented in linked data format in the Registry.
Partial translations covering the RDA Reference data of element sets and value vocabularies will also be added to the Registry.
The RDA Registry provides access to linked data versions of the RDA element sets and value vocabularies.
The Registry also provides access to tools for creating and editing RDA data and making it interoperate with non-RDA and non-FRBR applications.
RDA URIs are opaque and do not hint at their meaning or semantics. This avoids issues of language and changes of label.
RDA has developed lexical URIs that can be “read” more easily by developers. These can be in any language; the lexical URIs are automatically linked to the canonical URI.
Each RDA element set and value vocabulary can be freely downloaded in different versions or serializations of RDF.
Human-readable versions are available in the Open Metadata Registry
Allowing catalogers (and other data creators) to work with labels and guidelines has the potential to improve the quality of data available.
RDA: Alive and Well and Still Speaking MARC
A N D S T I L L S P E A K I N G M A R C
RDA: ALIVE AND WELL
RDA: MORE THAN RULES
• RDA started out as AACR3, but that was a long time
ago. Now RDA includes:
• The RDA Registry: full versioned data services for libraries,
vendors, and others using RDA elements and content
• A world-wide schedule of training efforts using RIMMF to show
how RDA improves access to cultural heritage materials (see
http://rballs.info for listing and data)
• The RDA Toolkit, a tool containing guidelines for principled
application of RDA for a variety of materials
• Multilingual capability for French, Spanish, Italian, Chinese, with
others to come …
• Includes instruction, vocabularies, data ...
LINKED DATA, FRBR AND RDF
• RDA data model based on FRBR (FRBR-less version
available as ‘unconstrained’)
• Using Resource Description Framework (RDF)
• Designed for efficient machine-processing of metadata at global
scale (Semantic Web)
• Based on ‘statements’ rather than ‘records’
• Statements from a variety of sources can be aggregated
(providing multiple points of view & curated ‘packages’ of data for
RDA IN RDF
• RDA controlled terminologies represented as RDF value
• Example: RDA carrier type vocabulary
• Entities, attributes, and relationships represented as
RDF element sets
• Entities represented in RDF as classes
• Attributes and relationships represented in RDF as properties
• Example: RDA manifestation properties
Multilingual linked data
RDF as global language
Formal or canonical URI is “opaque”
- not intended for human readability
Lexical URI is intended for human readability
in different languages
RDF linked data serializations
Except HTML for humans
Unconstrained linked data
Relates a resource to an agent
contributing to a resource.
German translationSwedish work English cataloguer
Linked data vs Semantic Web
All nodes identified
Easy to develop with
Hard to develop with
• Open Metadata Registry (OMR) = http://metadataregistry.org
• RDA Registry = http://RDAregistry.info
• RDA Registry issues reporting =
• Rballs.info = http://rballs.info
• RIMMF =
• Diane I. Hillmann (email@example.com)
Thanks to Gordon Dunsire for the use of his slides