I had the honor of being invited to participate in a linked data panel at the ALIA (Australian Library and Information Association) Information Online conference held in Sydney last month. We panelists were pleasantly surprised by the turnout—about 200 people attended our session, more than anticipated (standing room only). The conference organizers were amazingly prompt at uploading all presentations for the entire conference on the ALIA website.
Here’s my summary of the session’s highlights, reinforced by what was tweeted (#online17).
Monika Szunejko (National Library of Australia): “The links effect: The laws of attraction, linked data, and the national union catalogues of France and Britain” (Rather than speaker notes on the PowerPoint, see the accompanying paper.) Monika highlighted key findings from her case studies of the British Library’s and Bibliothèque nationale de France’s releases of their respective national bibliographies as linked data to inform the National Library of Australia’s own plans.
- The BL wanted to move away from library-specific formats and adopt standards that would reach audiences outside the library world; data must be open and re-used. It also wanted to enrich professional practice, leveraging staff skills in scripting languages and data conversion. Its success was affirmed by unlocking the value of its national bibliography to broader audiences, influencing Google rankings for its unique material and developing in-house expertise in linked data principles and data modeling. It also built an infrastructure that could be shared with other libraries.
- The BnF’s main impetus was to connect different databases, breaking down their silos—“a wedding planner between data sets”—enhancing discovery across collections and making them visible on the web to reach non-library audiences. It already had a deep engagement with identifiers as one of the founders of both VIAF and ISNI, and saw linked data as a means to expand their professional practices.
Both organizations see disseminating information as part of their mission, purpose and value and linked data helps them achieve that goal, while reaching new audiences and maximizing their data assets. Monika concluded that it was important for each library to understand its own context—your own reasons for moving to linked data—and your own capacity and capabilities. She encouraged others to join communities of practice to inform their own implementations.
Jenny Langenstrass & Katrina McAlpine (U. Sydney): “Good data bad data: getting ready for linked data” (the PowerPoint includes speaker notes). This presentation focused on the steps libraries could take now to prepare their legacy MARC data for future linked data implementations, such as data remediation and using identifiers. The University of Sydney has focused on authority work, and particularly disambiguating personal names (over 40 entries for “Robert Scott”). Research information management systems already use identifiers like DOIs and ORCIDs and those for grants so that researchers can be linked to their papers, institutions, grants, datasets and collaborators. Katrina pointed to the Research Data Switchboard (http://rd-switchboard.net/) as a successful example of linking researchers, organizations, datasets and papers across the data collections in Research Data Australia, Dryad, CERN and other international repositories.
Mine on “Linked data – bringing the world closer together” (PowerPoint includes speaker notes). My presentation focused on the opportunities of embedding library data into the “web of data”, not just the “web of documents” we’ve become familiar with. I focused on where a large portion of our communities live and learn—Google and Wikipedia—as platforms where we could leverage the work libraries have done to expose relationships among works, translations and their creators. I also showed some linked data examples from the 2015 International Linked Data Survey for Implementers to illustrate how library data could bridge domains and languages.
There were many good and inspiring sessions during the ALIA Information Online 2017 conference and I encourage you to look at the full program.
Karen Smith-Yoshimura, senior program officer, works on topics related to creating and managing metadata with a focus on large research libraries and multilingual requirements.