Linked data for libraries

While traditional library metadata is based on a proprietary, document based model, linked data uses a dynamic data based model based on linking simple three part RDF statements ( ‘triples’) describing resources (e.g. books, persons, corporate bodies, places, etc.). The subject and the predicate of the triple statement must be URIs but the object can be either a literal value or a URI. More recently RDF 1.1 has also introduced the IRI.

A move to linked data therefore is one from self-contained ‘document style’ bibliographic records describing published resources to one in which simple statements about resource (e.g. [This book] [has the author] [Charles Dickens]) are created with ‘records’ flexibly assembled  from selected statements as required by linking those statements sharing common properties.

Linked Open Data services are based on the use of large scale sets of RDF triple statements (‘triple stores’) which can be queried using the SPARQL standard and made accessible via a linked data API. Results can be delivered in a number of formats (e.g. JSON, Turtle or RDF).

Bibliography, Best practices

Last update: 5 December 2017


‹ Previous page
Next page ›