Participate in research to refine requirements

Linked data has the potential to offer libraries a way to make connections between items that isn’t possible through traditional cataloging standards. OCLC has been a leader in exploring how linked data can make libraries more interconnected and visible online. In 2017, OCLC began a pilot project that brought together metadata staff from a variety of library types with OCLC researchers, product managers, and software engineers for a hands-on experiment with a prototype designed to open linked data opportunities to libraries.

As one pilot project participant, Metadata Strategist Stephen Hearn, explained, “At the University of Minnesota, we really wanted to do something with a linked data model, and this was a relatively easy way to get into that.” Not only did the pilot project provide that opportunity, but it was also fun to participate in. “What I like best about the prototype is that it’s not focused primarily on bib records, it’s on describing bib objects,” he continued. Principal Cataloger Sarah Ross from Cornell University added, “The prototype is beautiful. It’s intuitive. And I like the idea of introducing relationships and making them available to users of the catalog.”

Participants also recognized the flexibility of linked data to expand resource description beyond the cataloging team. “A lot of us have experience working with curators on special cataloging projects,” said Chew Chiat Naun, Head of Metadata Creation at Harvard Library. “But open linked data platforms like this one could potentially make metadata work a shared effort between curation and cataloging.” Sarah agreed, “You can imagine somebody who's not a cataloger but who does have rich expertise in a particular area being able to understand the context of an item in way that a cataloger wouldn't.”

One benefit of the pilot project was the close working relationships the libraries developed with OCLC. “It’s much more of a fruitful, two-way dialogue than we’re used to having,” Naun said. Stephen agreed, “The kind of expertise that the developers bring is unusual. It’s been really nice having the extensively capable help from the developers that we’ve had with OCLC.” And Sarah noted, “It’s a big help just to work with somebody you don’t have to explain cataloging to.” But beyond that, “I like the way OCLC tried something and told us when it wasn’t working and included us in finding a different way to do it,” she said. Stephen added, “OCLC Research in particular is made up of people who really understand what we do and care about it. And we work collectively with them. That’s a great thing.”

At the end of the pilot project, Stephen said, “We understand better now what linked data looks like when applied to the kinds of description we want to do.” And collaboratively working through the intellectual challenges of linked data was enjoyable too. “A lot of the fun in this project has been the ability to talk with people about specific problems in descriptive cataloging,” Naun said. Hearing from others, “it suddenly occurs to you that there might be a better way to do something.” Sarah shared an experience from a call during which Stephen presented a description of a poster advertising a concert. Sarah said, “The kernel was the event, not the poster—not the actual thing I’m cataloging, but the thing behind that. And this prototype can separate the two concepts and link them. I really like that.”

With the lessons learned from this pilot project, OCLC continues to work on designing a solution to help libraries use linked data to establish relationships in their catalogs in a way that can be recognized by searchers outside the library community. “We need something like this prototype built on a platform that is used more widely than just in libraries,” Naun said. “We need a shared community practice that isn’t as restrictive and gives us more flexibility in what we can say about the entities and resources we’re interested in.”

Project at a glance

• Partnered libraries with OCLC researchers, product managers, and developers to pilot a service showing the value of linked data in the library resource description workflow
• Allowed metadata specialists to test a prototype of a reconciliation service to find and disambiguate linked data entities and an editor service to view, create, and edit linked data descriptions and relationships
• Drew participation from 16 libraries, including public, national, and special libraries alongside academic and research libraries