Best Practices for Library Linked Open Data (LOD) Publication

Linked Open Data (LOD) is becoming an increasingly popular way of publishing data for others to use. The principle behind it is very simple. Firstly, there are different types of information and concepts, as well as the relationships between them. In this context, we will refer to these three components as ‘things’. When you publish data we, 1) use identifiers for all these ‘things’, 2) make the identifiers functional so they can be used to access these ‘things’. And, 3) provide useful metadata about the ‘things’ when you access them through their identifiers. As the term ‘linked’ implies, the ‘things’ from multiple resources can be linked to each otherresulting in an interconnected web of information which can be easily machine-processable. When the result is published for external and free use (using an open license), the result is LOD1 .

This document looks at publishing LOD from a library perspective and argues why it should be employed and how. We will not delve into technical details nor their respective technical tools. Instead, we will present various aspects of the topic, introduce different options available, and lay out a foundation for possible exploration at a later stage. As such, the body of this document presents the six steps of LOD publication and explains each one of these steps in depth. Thereafter, suggested readings are provided to help you delve deeper into the subject, if needed.