Co/Wy Service Hub meeting with DPLA Staff, Meeting Notes, September 2016

Members of the Digital Collections Planning Group (DCPG), the Metadata Working Group, the Technology Working Group and Digital Public Library of America (DPLA) staff met in September to review the Co/Wy Service Hub initiative.  The following areas were discussed:

Overview of the DPLA and its potential

Kelcy Shepherd, DPLA’s Network Manager, gave an overview of the DPLA.  DPLA currently has 14 million records and 15 content hubs and 14 service hubs in addition to three new hubs that were recently approved.  She shared information on the DPLA platform which allows for data sharing and gives free public access to information about digital collections held by libraries, archives and museums.  The DPLA provides institutions with the potential to create educational sets, create online exhibits, build a community around individual collections, develop a platform for underrepresented groups, and be on the same playing field as other organizations.  Joining the DPLA has increased traffic to collections, some by 50%. 

Service Hub Requirements and DPLA Expectations

Emily GoreDPLA’s Director of Content presented on the service hub requirements and DPLA’s expectations.

  • In the first feed the hub agrees to provide at least 50,000 records in one data format and in one feed. Preferably this format is DPLA’s  MAPV4, but otherwise Qualified Dublin Core works the best.
  • All metadata is free of copyright restrictions.
  • Metadata should have thumbnail links to provided content. Ideally the hub should use IIIF (International Image Interoperability Framework) to share images with DPLA .
  • The hub agrees to work on rights labeling and to use rightsstatements.org standardized rights statements.  DPLA will work with the hub to help fund local rights implementation training.

DPLA is currently:

  • Working on rolling out a membership model. There will be one fee per hub. Hubs in turn will have a governance voice and 2 seats on the board. The fees will make up about 20% of the DPLA operating costs.
  • Interested in partnering with content experts to build and curate collections.
  • Working to improve non-item level content access and working through a newspaper planning grant.

Emily also outlined DPLA’s workflow for working with new hubs.

  1. Submit application (must have at least 50,000 records)
  2. Approve application
  3. Introductory Webinar
  4. Data exchange agreement
  5. Ingest information form filled out- technical setup, feed info, data format.
  6. Initial metadata review (1:1 with DPLA staff).  This is an iterative process.
  7. Write scripts
  8. Ingest data into quality control environment. Perform QA.
  9. We update/review data
  10. Approve data in the dev environment
  11. Move to production
  12. Schedule recurring ingests based on how the hub expects to add content.

Emily shared examples of other hub models including South Carolina, New York (Empire State), Pennsylvania, North Carolina and Mountain West Digital Library. While there is no one-size-fits-all hub model, there is much that can be learned and adapted from other hubs.

  • South Carolina – Three-region system with loose MOUs among the three organizing entities
    • Clemson – Grant administration and regionally hosting collections
    • College of Charleston – Metadata, harvesting technology and regionally hosting collections
    • Univ. of South Carolina – Program coordination, community engagement, state grant administration and  regionally hosting collections
  • North Carolina – Managing institutions and roles are:
    • State Library of NC – Funding and staff
    • UNC Libraries – Administration, funding and staff
    • Digital NC – Program coordination, metadata remediation, technology, community engagement, hosted collections and aggregates collections.  This was an existing digital library before joining DPLA
  • Mountain West Digital Library
    • Multi-state aggregation
    • Have sub-hubs that offer hosting and many offer digitization services
    • Has membership fees and a setup fee. These fees help pay for staff
    • Montana Memory Project and University of Montana recently left Mountain West to create a Montana service hub.  The MW membership fee is per institution not per state, therefore it was more fiscally responsible for Montana to create their own hub and pay one fee to DPLA

Data Harvesting at DPLA 

 Gretchen Gueguen, DPLA’s Data Services Coordinator, shared DPLA’s steps for harvesting hub data.  

  1. Analyze feed
  2. Mapping Crosswalk
  3. Mapping Code
  4. Harvest and Map
  5. Enrichment (DPLA’s version of the record)
  6. Quality Assurance
  7. Test portal
  8. Data into DPLA

The most basic elements required by DPLA are title, rights statement and URL.

Brainstorm session for Co/Wy Hub Model

The group then discussed the proposed Co/Wy hub model (co-wy-service-hub-model). This model outlines the proposed responsibilities of  the Colorado State Library, the partner organizations and the continued work of the initial planning groups and eventual governance group.  We proposed to do the hub work in three initial phases.  Phase one will include six organizations that will help the hub determine processes and workflows.  The hub will add more partners in phases two and three.  With this model, DPLA staff feel that the DCPG is ready to submit a hub application.  The Colorado State Library will draft the application to be reviewed by the DCPG.  The application will be submitted prior to January 1.

If you have any questions about the Colorado and Wyoming DPLA Service Hub initiative please contact me at, ljeremias@coloradovirtuallibrary.org.