2025-04-23

Housekeeping

Action items from last meeting:

• Everyone agrees this is a good approach, since the Metadata Basics LibGuide is most likely to be kept up to date.

• Melanie has been experimenting with using a macro (excerpt/insert macros) on the wiki to automatically add resources links shared across our meeting notes to the Resources page. It is a little bit clunkier that Melanie had expected but more straight forward than copy/paste.

Yes:

• Author with LCNAF, no wikidata entry → minimal with placeholder icon for photo

• Author with LCNAF and Wikidata entry without LC ID property → minimal with placeholder icon for photo

• Author with LCNAF and Wikidata entry with LC ID property → best case scenario

Takeaway: LC ID (P244) property is required in Wikidata entry for Wikidata/Wikipedia linkage & display to work in Primo

Announcements:

• Not today

Updates from the field

• Posted to ld-learning list this morning

  • To inform an ELUNA presentation (Navigating Challenges and Embracing Opportunities: Advancing Linked Open Data Initiatives in Libraries), presenters would like to incorporate some information gathered from other Alma institutions. They are conducting a brief survey to gather information on the various frameworks and initiatives employed by libraries to promote and advance the use of linked data.

• ELUNA/IGeLU Linked Open Data Community of Practice Working Group hosted a Getting Started with LC BIBFRAME in Alma meeting on April 22nd. Recording pending, but there were several resources shared ahead of the meeting (below). Melanie’s notes from meeting.

  • For getting the most out of the session, we encourage attendees to review the following resources before the meeting:

    • “Using the Ex Libris API Console: Adding, Updating, and Deleting BIBFRAME records “by the each” (Healther Pretty, Memorial University of Newfoundland) -- [video, 22 minutes]

      • Using the Ex Libris API Console (slides)

    • “Discussion on Importing BIBFRAME with the Alma API Jupyter Notebook” (Jim Hahn, University of Pennsylvania)  -- [video, 15 minutes]

    • “Importing BIBFRAME into Alma with Python” (Stefan Schuh, Austrian Library Network and Service Ltd.) -- [Python code on GitHub]

  • Additionally, if you’d like to explore the features mentioned in the resources above in your own Alma environment, you'll need to:

    • Enable BIBFRAME features following instructions outlined in the “Enabling BIBFRAME Features” document

    • Set up an API Key (instructions are in this slide presentation)

  • Questions for Discussion Google doc

• Ann – Cataloguers at UT Libraries are not currently doing any work with BIBFRAME in Alma. Most likely they will wait to see how other libraries implement BIBFRAME within their workflows before moving forward with it.

Open data - presentation and discussion

Two guest speakers from UT Libraries: Michael Shensky, Head of Research Data Services, and Bryan Gee, Open Research Coordinator for Data and Software

• Slides

• Research Data Services focus is obviously research data so that is the focus of the presentation today

• Open data in a research context - core 4 principles (FAIR)

  • Findable (DOIs, etc…)

  • Accessible (permissions, file formats…)

  • Interoperable (standards, controlled vocs…)

  • Reusable (metadata about the metadata)

• Types of open data

  • Dataset from a research data repository

  • Data from a local/state/federal government portal

  • Data shared by companies

  • Data shared by non-profits and NGOs

• Examples of open data

  • Global Biodiversity Information Facility (GBIF)

• What is Linked Data?

  • URIs

  • HTTP

  • Dereference to useful information

  • Links to other URIs

• What is Linked Open Data?

  • Five star model as defined by Tim Berners Lee

    • Available in the web and with open license

    • Machine readable

    • Non-proprietary format

    • Open W3C standards (RDF ad SPARQL)

    • Linked to other people’s data to provide context

    • Linked data in practice

  • Requirements for being added to Linked Open Data Cloud

    • Resolvable http:// (or https://) URIs

    • Resolved to RDF format (with or without content negotiation

    • Dataset must contain at least 1000 triples

    • Data set must be connected via RDF links to at least 50 links

    • Access to the entire dataset should be through

• Open data and linked data

  • Data can be open and not linked to other data

    • Not referencing to anything external

    • They might reference entities also referenced in other datasets but not using URIs

  • Data can be linked but not open

    • RDF, URIs, HTTP protocol, SPARQL

• Open linked Research Data

  • RDS is trying to figure out how they can link UT research output to other datasets

    • Bryan has been focusing in UT research data

    • Michael has been trying to find UT developed research software

  • Current approach

    • Use of PIDs for individuals (ORCID) and organizations (ROR, GRID) – connects to additional research output

    • Explicit links between related objects (dataset, software, preprint article…)

    • Consistent use of controlled vocabulary and metadata schemas

    • Minimum metadata standards

  • However, right now they are not publishing their data as RDF

    • Right now using JSON and APIs

• Questions for the group

  • Any ideas of what RDS do with LOD?

  • How can the RDS team help?

User study on the Primo Person card and page - presentation and discussion

Guest speaker: HRC FWS employee and UT iSchool student Irene Sibi

• Slides

• Known issues

  • Not all records display contributor cards because not all of them are in wikidata

  • Not all author cards include the same amount of data

• 7 faculty and researchers (people very familiar with the catalog)

• 1 hour in-depth interviews with 2 different scenarios (known author with a fully fleshed author card and author with a less populated author card)

• Has put together an affinity diagram including groupings of issues and quotes by the individuals pin pointing those issues

• Main issues and take aways identified

  • No way to tell how the people associated are associated

  • Information might be inconsistent (instances where the authorship does not match)

  • Context is very helpful to people (give the user all the autonomy they need to know what to do with the information)

  • Concern about loosing catalog data trust based on those inconsistencies, or how those relationships have been stablished

  • Concern about authors being more highlighted than others based on the amount of data in their author card

  • Issues with loading and navigation

• Question

  • Ann – did any of the interviewees brought up privacy concerns?

    • Did not come up in the interviews, but it is a good point. Danger of merging personal data from wikidata with professional data from an academic environment (two different environments)

• Irene will be putting together a report and will share with the group

• Potential to bring these issues to ExLibris

Wrap-up

• Both presentation slides will be shared in the wiki

• Next meeting we will debrief as a group