The Fast & the FRDR Tuesday June 11th, 2019 Alex Garnett, Simon - - PowerPoint PPT Presentation

The Fast & the FRDR

Alex Garnett, Simon Fraser University, garnett@sfu.ca; Lee Wilson, Portage, ACENET, lee.wilson@ace- net.ca; Clara Turp, McGill University, clara.turp@mcgill.ca; Julienne Pascoe, Library & Archives Canada, julienne.pascoe@canadiana.ca

Tuesday June 11th, 2019

Agenda

• Background information
• Discovery and Metadata
• The Fast & the FRDR

Background

The Portage Network

• The Portage Network:
• Canadian Association of Research Libraries (CARL)
• Shared stewardship of research data in Canada
• Networks of experts and working groups

Portage Expert Groups

• Data Management Planning
• Curation
• Data Discovery
• Preservation
• Training
• Research Intelligence
• Data Repositories

Data Discovery Expert Group

“The main purpose of the Data Discovery Expert Group (DDEG) is to support data creators and curators in planning, producing, and managing descriptive metadata to enable the effective discovery and reuse of research data across a wide range of disciplines.”

FRDR Discovery Service Working Group

• Federated Research Data Repository
• Partnership with Compute Canada
• 100 000 metadata records from over 40 repositories
• https://www.frdr.ca/repo/
• FRDR Discovery Service Working Group
• Appropriate description and discoverability
• Discoverability for the widest audience

Discovery & Metadata

Discovery

• Discovery: research data to support new research
• The role of metadata
• FRDR as aggregation service

Harvested metadata

• Challenges in normalizing metadata
• Crosswalks built for standard fields
• Subject/Keywords

The Fast & the FRDR

FAST

• "Faceted Application of Subject Terminology"
• Derived from Library of Congress Subject Headings
• Maintained by OCLC
• Why did we choose FAST?
• Fairly widely implemented with a mature basis
• Sufficiently generic, wide-reach
• Good open APIs, potential for linked data

Open Refine

• Open source Google project
• Powerful for this use case:

○

Messy dataset

○

Reconciliation service using open APIs

○

Using a semi-mediated process https://github.com/cmharlow/fast-reconcile https://github.com/OpenRefine/OpenRefine/wiki/Reconciliation

Challenges & Next Steps

Challenges

• Large-Scale and ongoing work
• What about unmapped keywords?
• Integrating this work with the FRDR harvesting process?

Fast & FRDR W orking Group

• Define guidelines for mapping subject keywords that do not

correspond directly with the FAST vocabulary

• Work collaboratively and asynchronously using OpenRefine to map

subject keywords to FAST vocabulary

• Peer review mappings to ensure consistency and quality
• Develop a work plan for unmapped subject keywords
• Develop a work plan to ingest those mappings back into the

repository and workflows to map new metadata harvests.

Discussion Points

• The Bilingual Canadian context

○

Mapping French keywords (RVM)

• Choosing policies and practices:

○

Granularity: Royal canadian mounted police superannuaction account

○

Standardization

○

Abbreviations, names, numbers, multivalued data

Questions?

Thank you!