Complex topics and tough problems are better tackled in a group. One way to begin to look more closely at the practical details of linked data is to form a group of interested colleagues and start reading and discussing together. At UNLV we formed a group with representatives from many functional areas of the library: Digital Collections was strongly represented but we also had members from Technical Services, Systems/Web, the IR department, and Assessment. Not everyone had the same technical expertise, but we found that by creating an atmosphere where all questions were welcome that the group became a useful venue for working together to understand the basics of linked data. The group was led by two librarians who often assigned readings or background material for the agenda, but no one on the group started as a linked data "expert", we all learned together. Sometimes it was a little chaotic. Sometimes the materials we chose were way above our heads, or so technical that were all bored to death. But other times we had little breakthroughs and these generated momentum for the group..Below are some of the topics we tackled in our first year.
- What is linked data?
- Overview of linked data projects and potential use in library settings
- Basic Concepts (URI, RDF,statements/triples -- subject, predicate, object)
It was at this point when we decided to try to develop a linked data project transforming digital collections metadata into linked data. The study group branched a bit into a group more interested in MARC and BIBFRAME and the digital collections team focused on proposing a project. The project was further defined and this became the more intense focus with the UNLV Linked Data Project Team providing regular updates to the study group on topics such as explorations into:
- RDF and linked data technologies (OpenRefine, Mulgara, Virtuoso)
- General use data elements; ontologies and controlled vocabularies to be used as values for data elements
- SPARQL query language
Information was shared widely in the study group and our library to grow support for the project. The work of the study group and the project has been documented in an internal wiki with comprehensive listings of resources, reference materials, meeting minutes, and presentations. After 18 months from the initiation of the study group, a demonstration of creating linked data with digital collections metadata had been completed. This work was completed by two busy librarians (with full-time job responsibilities) who were assisted by one Graduate Assistant with a background in MIS. All of the work on our project was done through questioning, researching, experimenting, testing, and revising. There was no recipe and no manual to follow. We relied heavily on those in the linked open data community for resources and inspiration. Of particular note was the excitement we felt in discovering LODLAM. It was a good feeling to finally discover other people working on practical linked data projects.
We share this information to encourage others to move forward and explore. You do not need to be a programmer or understand everything about the technology to create linked data. It can begin with scribbling down a few questions and gathering some colleagues together to brainstorm ideas. Who knows where this modest beginning might take you?