is a gateway to raw RDF data from Oslo Public Library and a platform for developers to access these data and build applications that use them.

The main part of is an enriched RDF version of the library catalogue and its authority files. This includes data about books, music, movies, persons, organizations, topics and genres. The dataset consists of an RDF version of the main MARC catalogue, enriched with book cover URLs harvested from external datasets, as well as a FRBR-like work-manifestation-structure that identifies the unique works described in the catalogue, and clusters the different editions and translations of each work. The dataset is updated daily to account for additions, deletions and changes made in the catalogue. also contains a collection of book reviews produced by librarians at Oslo Public Library and other libraries from all over Norway. Some of them through the former «Ønskebok» project. Book reviews are linked to works in the main catalogue dataset.

As of january 2014, contains descriptions of:

~ 415 000 documents (books, CDs, DVDs, periodicals etc.)
~ 200 000 persons
~ 75 000 organizations
~ 65 000 topics
~ 2000 genres
~ 7500 book reviews

To look up data in the web browser, simply enter a resource URI in the address bar:


As Turtle:




The graph names listed below are not web addresses that locate web resources; they merely serve as identifiers (URIs). You will not find anything meaningful if you paste the graph names into a browser’s address bar. To look up data from the graphs you must use SPARQL queries through the endpoint we link to from this page. Several example queries are listed at the bottom of the page, if you want to experiment.

• Enriched RDF version of the library catalogue and its authority files.

• Bookreviews from Oslo public library and the «Ønskebok» project

• A set of statements about assumed similarity between literary works. Based on subjective evaluations as well as automated methods.


The data

Descriptions of the library’s books, music and movies make up the main part of the dataset. The documents are mainly described through predicates from the Dublin Core, Bibo and Fabio vocabularies, but these are supplemented with other predicates where needed. The metadata includes almost all the information given in the original catalogue records, in addition to data from external sources, such as links to cover images, book reviews and digitized texts.
Each document in the collection is declared to be the manifestation of a work. This can be used to cluster the different translations and editions of a work, or to identify the unique works in a set. Work descriptions contain ISBNs from all manifestations of the work, in addition to ISBNs harvested from external sources, so it is possible to look up a book by ISBN even if that particular edition is not part of the library collection.
Persons are mainly described by use of the FOAF vocabulary. Information about persons include names, nationalities and years of birth and death.
Subjects are described with the use of the SKOS vocabulary. They are labeled with strings of subject headings that contain both the main topical term as well as aspects about the topic. Each subject is linked hierarchically to a constructed subject, based on the main topical term alone, so that related subjects are clustered.

<> a skos:Concept ;          (Main topical term: Chess)
skos:narrower <> ;    (Chess – Periodicals)
skos:narrower <> ;    (Chess – Endgame studies)
skos:narrower <> ;    (Chess – Problems)
skos:narrower <> .    (Chess – History)
Book reviews constitute a separate graph in the RDF triple store. They are described mainly with the Review vocabulary. In most cases reviews describe works, but there might also be cases where a review describes the language in a particular translation or the illustrations in a particular edition. Therefore reviews are linked to from both the work it describes and the manifestation it is based on.


SPARQL endpoint

To query the datasets, use the SPARQL endpoint:

The endpoint returns data in several formats, including RDF/XML, JSON and n3/Turtle.

Example queries

#Returns all statements about the resource with that URI:

<> ?p ?o .

#Returns the name of the creator of the resource:

PREFIX dct: <>
PREFIX foaf: <>
<> dct:creator ?creator .
?creator foaf:name ?name .

#Returns URI and title of all documents by the same creator, sorted alphabetically by title:

PREFIX bibo: <>
PREFIX dct: <>
SELECT ?document ?title WHERE {
<> dct:creator ?creator .
?document a bibo:Document ;
dct:creator ?creator;
dct:title ?title .
ORDER BY ?title



Bibliographic data (data from the graph <>) is licensed under CC Attribution-NonCommercial-ShareAlike.
Book reviews (data from the graph <>) are licensed under CC Attribution-ShareAlike

Working code

Right now we have two applications that use the datasets

  1. Active shelves (Aktive hyller): Code on GitHub
    The Active shelf is a physical touchscreen device that makes use of open source software, RFID technology, RDF data and external web service APIs to provide information about any library book a patron is curious to know more about. See video.
  2. Book reviews database (Bokanbefalinger): Code on GitHub
    The Book reviews database collect book reviews from Norwegian libraries in one database. The reviews are stored as open linked data in the library’s RDF store. The reviews are described with metadata, and linked to books in the linked data version of the library catalogue. Webpage Book reviews.

Om forfatteren