Organization Name Linked Data


Alphabetical Index


Overview

The NCSU Organization Name Linked Data (ONLD) is based on the NCSU Organization Name Authority, a tool maintained by the Acquisitions & Discovery department since 2009 to manage the variant forms of name for journal and e-resource publishers, providers, and vendors in E-Matrix, our locally-developed electronic resource management system (ERMS). The Organization Name Authority was first described by Kristen Blake (Wilson) & Jacquie Samples in "Creating organization name authority within an electronic resources management system," Library Resources and Technical Resources, 53(2), 2009, p. 94-97, which is available online here.

The names chosen as the authorized form reflect an acquisitions, rather than bibliographic, orientation. For example, in the Library of Congress Name Authority File, the Institute of Electrical and Electronics Engineers is represented by the full name, where as in the NCSU ONLD, it appears as "IEEE," which is how it is generally known among acquisitions staff. Also, there are many subsidiary units with valid headings in the LC Name Authority File but for the purpose of managing journals and electronic resources they are simply considered to be variant forms of names for the parent organization that manages acquisitions and licensing-related functions for the subsidiaries.

The information in the NCSU Organization Name Linked Data are represented as RDF triples using properties from the SKOS, RDF Schema, FOAF, and OWL vocabularies. Clicking on the name of each property will take users to the property's definition. The authorized form of name for each organization is recorded with skos:prefLabel and variant forms of name were recorded with skos:altLabel. All of the organizations are associated with relevant classes in several popular vocabularies using rdf:type. The webpage of the organization is recorded using foaf:homepage.

Where possible, Acquisitions & Discovery staff created links to descriptions of the same organization in other linked data sources, including the Virtual International Authority File (VIAF), the Library of Congress Name Authority File (LCNAF), Dbpedia, Freebase, and International Standard Name Identifier (ISNI). These links will enable users of the data to easily incorporate properties from other linked data sources in future applications and are encouraged in Tim Berners-Lee's description of 5 Star Open Data. The links were initially recorded using the owl:sameAs property but in November 2014, we chose to switch to the skos:closeMatch property after becoming aware of the ongoing debate about the overuse of owl:sameAs, which has very strong semantic implications. Halpin, Hayes, McCusker, McGuinness, & Thompson’s 2010 article "When owl:sameas isn’t the same: an analysis of identity in linked data" serves as an excellent summary of this issue.

The data is available in RDF-XML, N-Triples, N3/Turtle, and JSON-LD. RDF triples can also be extracted from the RDFa metadata embedded in the HTML page for for each organization.

This data set will also be the seed data for organizations in the Global Open Knowledgebase (GOKb), a freely available data repository with key publication information about electronic resources that will have its public release in September. As a part of NCSU’s lead role in the GOKb project, we are collaborating with the GOKb developers on future linked data initiatives involving title, package and platform data.

This linked data is made freely available with the Creative Commons CC0 License.


Downloads

Download the entire NCSU Organization Name Linked Data set in different formats. Data last updated 12/22/15.


View Sample Pages

Here are two organization pages with the full range of properties available in this data set.


Contact

For questions or reporting broken or incorrect links, please contact:

Jacob Shelby

GOKb Editor

Acquisitions & Discovery

NC State University Libraries


Sample XML and XSLT Files

For anyone interested in how this linked data set was created, here is a sample XML file of NCSU ONLD data and the XSLT files that were used to produce various RDF serializations.

XML

XSLT


Tools Used

  • Oxygen XML Editor 15.2 - XML editor used to create XSLT files to convert data into N-Triples and RDFa-embedded HTML.
  • Open Refine - Formerly Google Refine, an open source data cleanup and transformation tool. The RDF Refine extension was used to test batch searching for relevant URIs from DBpedia.
  • Sesame 2.7.11 - Open source triple store used for experimenting with conversions to various RDF serializations.

Organization Name URI Look Up Script

During the initial development of this data set, Acquisitions & Discovery staff conducted manual searches of other linked data sets to find matching URIs that could be added as links using the skos:closeMatch property. This was a time-consuming process and we later experimented with various ways of conducting automated searches of other linked data sets for matching URIs. The most effective approach was the LCNAF Named Entity Reconciliation script created by Matt Carruthers at the University of Michigan Library. The ONLD project team modified this script to search for VIAF, LCNAF, ISNI and DBpedia identifiers for the organizations in our data set.

Below are links to the Organization Name URI Look Up Script and a sample set of organizations to load into OpenRefine as a project. To run the script, navigate to the "Undo / Redo" tab, click on the "Apply" button and paste the script into the text box and click the "Perform Operations" button. Running the script on this list should take approximately 9 minutes.


Publications

  • Hanson, Eric. A Beginner's Guide to Creating Library Linked Data: Lessons from NCSU's Organization Name Linked Data Project. Serials Review. Volume 40, Issue 4. DOI: 10.1080/00987913.2014.975887.

Presentations


Last Updated

10/01/2019


License

creative commons license public domain
To the extent possible under law, NC State University Libraries has waived all copyright and related or neighboring rights to NCSU Organization Name Linked Data. This work is published from: United States.