Applied Linked Data 2017-04-17

Owned by Bill Branan
Jun 16, 2017
2 min read
Time: 12:00 PM PT / 3:00 PM ET
Backup Phone Line If Hangout Doesn't Work: 712-775-7035 (Access Code: 960009)


Attendees:
    
Agenda:
  1. Brief introductions
  2. How often will we meet?
    1. Once per month, 3rd Monday 3pm Eastern
  3. Notes: continue using etherpad?
    1. Executive decision to just use wiki
  4. Revised scope / objectives
    1. We reviewed
  5. General discussion driven by questions about technical architecture and infrastructure from Rutgers attendees
    1. They will post some documents on the wiki and for some feedback / review next time 
    2. Grant will build on hyrax
      1. document several kinds of resources in Fedora
        1. research data
        2. intermediate data from tools
        3. people (may link to ORCID)
      2. pull into an active discovery layer OAI/ORE
        1. relationships between people, tools, and intermediate products resulting from that analysis
        2. create a resource map
        3. interdisciplinary in nautre: this means it has to provide lots of context so different people can understand the data
      3. decided not to use VIVO; they were told it was overkill for this project
      4. research data / building blocks grant (co-PIs with CS department)
      5. Another part of the grant is about building a repository to be used across Pennsylvania (penn state, temple, drexel) and New Jersey
      6. But this portion is mostly a virtual repository for linking resources.
  6. Ryan Wick's overview of current work at OD: 
    1. can share what Oregon Digital has done with linked data

    2. local authorities with ControlledVocabularyManager powering opaquenamespace.org

    3. interested in new developments and other's integrations with Questioning Authority and LDF

    4. OD2 dev looking at starting with Hyrax, expanding interfaces for and management of linked data

    5. OSU is currently migrating from DSpace and currently expanding linked data in our IR: Scholars Archive (building on Hyrax 2)

  7. Ryan Johnson current goings on at UC San Diego:
    1. We currently have a Hydra app with just triplestores, no Fedora. We manage all of our local authorities (thousands, mostly topics and names), which then express matches to external authorities (VIAF, LoC, Wikidata etc.). We did a large reconciliation using mostly basic text-breaking scripts (many topics were complex and pre-coordinated strings) and reconciliation tools (OpenRefine recon scripts, scripts using SPARQL)
    2. We are now moving to Hyrax (soon?) so obviously questioning authority is now very relevant to us, but we still will try to keep the thousands of local authorities, and undergo ongoing reconciliation work
    3. We also have a tool to import batches of new authorities, and which will be re-done in Hyrax. We do this before we ingest new collections to production and have reconciled the authorities. Is anyone in the community interested in this kind of thing being gem-ified? 
    4. We already have many Research Data collections and so are interested in what other people are looking at here in terms of linked data authorities. I've found lots of great linked data authorities in Wikidata, but would like to explore the kind of data we could cache there
  8. Review action items from LDCX meeting
  9. Please add!


Next Meeting:  May 15 at 12pst/3est