Library of Congress

Authorities & Vocabularies

The Library of Congress > Linked Data Service > Technical Center: Metadata

Metadata

Metadata structure standards used by the Linked Data Service include:

When the Linked Data Service was first launched in 2009, then called the Authorities and Vocabularies service, the data model used was influenced entirely by SKOS. As a result, most of the metadata associated with our offerings are disseminated using properties and classes from SKOS.

In 2010, LC developed MADS/RDF, which is a more specifically defined data model to represent the complexities of authority, vocabulary, and thesauri data than possible with SKOS alone.

MADS and MADS/RDF

The Metadata Authority Description Schema (MADS) is an XML schema for an element set that may be used to provide metadata about authorized forms of agents (people, organizations), events, and terms (topics, geographics, genres, etc.). The terms may be part of thesauri, taxonomies, subject heading systems, or other controlled value lists. MADS serves as a companion to the Metadata Object Description Schema (MODS) to provide metadata about the authoritative entities used in MODS descriptions.

MADS/RDF builds on MADS/XML as a knowledge organization system.  It is closely related to SKOS, the Simple Knowledge Organization System and a widely supported and adopted RDF vocabulary.  Given the close relationship between the aim of MADS/RDF and the aim of SKOS, the MADS ontology has been fully mapped to SKOS.

Unlike SKOS, however, which is very broad in its application, MADS/RDF is designed specifically to support authority data as used by and needed in the LIS community and its technology systems.  For example, MADS/RDF provides a means to record data from the Machine Readable Cataloging (MARC 21) Authorities format in RDF for use in semantic applications and Linked Data projects. MADS/RDF allows for more precise identification of parts of terms and better interoperability with LIS data.

Learn more about SKOS and RDF

Start with the SKOS Primer Offsite link, and then have a look at the SKOS Reference Offsite link. These contain a good overview of the basics of RDF, as well.

Additionally, have a look at this DevX article Offsite link, which describes the application of SKOS ConceptSchemes, Concepts, and Collections.

Who else uses SKOS?

As part of the TELplus Offsite link project, Antoine Isaac Offsite link offers a service Offsite link that uses SKOS to provide RAMEAU Offsite link subject headings as Linked Data Offsite link. In fact, in the spirit of Linked Data, this TELplus project and the Authorities and Vocabularies service link in kind between concepts that possess a semantic relation. This is made possible thanks to 60,000 mappings manually provided by the MACS Offsite link project.

OCLC is building tools with SKOS, and the United States National Agriculture Library Offsite link has released its thesaurus in SKOS. The National Science Digital Library Registry Offsite link also uses SKOS as a primary metadata model.

The United Nations Food and Agriculture Organization Offsite link uses SKOS Offsite link within its GeoNetwork Offsite link application.

Why SKOS?

SKOS is quickly becoming a prevalent standards-based tool for representing thesaurus data, as witnessed by the user list above. Using SKOS instead of creating a custom XML Schema approach minimizes reinvention of the wheel. Implementing custom XML Schemata often requires new, custom tools to be created by each user, whereas SKOS and RDF have numerous tools to make instant use of the data.

Furthermore, the notion of dereferenceable, REST Offsite linkful URIs is a strength of the Semantic Web and Linked Data community approach to data -- one which is attractive enough for us to leverage. It is not necessarily an approach that the Semantic Web community invented, nor is it something that XML is devoid of, comparatively speaking. However, centering services or content negotiation around a reusable or multi-purposed URI has not been a common design principle in most XML applications.

Lastly, the ability to see our data tied into the Linked Data world is attractive. Linked Data is heavily driven by Semantic Web and RDF technologies. This offers an exciting opportunity to bring our standards offerings to a new user community.