Finding Aid-LD: Implementing Linked Data in a Finding Aid Environment

Jacob Shelby

@jacob_shelby421

Overview

  • Archival Experience as Linked Data
  • Linked Data Publishing Strategies
  • Proposed Framework for Publishing Finding Aids as Linked Data
  • Proof-of-Concept Application of Linked Data Finding Aids
  • Advantages of Linked Data
  • Challenges
  • Steps That Can Be Taken to Prepare for Linked Data in a Finding Aid Environment
  • Future Research

Archival Experience as Linked Data

  • The archive world deals in hierarchy
  • Current practice is with EAD/XML

Collection

    Series

        File

            in box 2, folder 10

        File

            in box 2, folder 11

    Series

        Subseries

            File

                in box 10, folder 1

Archival Experience as Linked Data

  • Linked Data hierarchy comes from linking + things

Collection

Series 2

Series 1

Box 10

Box 2

Subseries 1

File 3

File 2

File 1

Folder 11

Folder 10

Folder 1

Publishing Strategies: Data Dump

  • Large, zipped file of a provider's full dataset
  • Use-case: User wants to download full dataset
  • Example: DBpedia data dump

image from DPLA bulk download

Publishing Strategies: Triplestore and SPARQL

  • Triplestore is a database for Linked Data
  • SPARQL is the query language for RDF
  • Use-case: Researcher wants to run complex, semantic querying of the data
  • Example: DBpedia SPARQL endpoint

Publishing Strategies: Triple Pattern Fragments

Publishing Strategies: Subject Pages

  • Creation of a document for each resource (similar to a traditional metadata record)
  • Content negotiation
    • URI redirects to human or machine readable file based on HTTP ACCEPT
  • Use-case: User wants to dereference a URI and discover new knowledge by browsing links
  • Example: DBpedia subject page for Christopher Lee

Proposed Method for Publishing Finding Aids

Subject page for each resource

  • e.g. Collection, Series, Box, Item, *Finding Aid
  • Content negotiation, HTML and RDF
  • Embedded metadata in the HTML
  • Flat folder structure
  • Arbitrary URIs for resources

Proposed Method for Publishing Finding Aids

Proof of Concept

Advantages of Linked Data

  • Semantic data = more meaningful querying
  • Strong framework for linking to and among resources
  • Modular data
  • Varied and specialized ways of publishing and presenting data
  • Straightens out "complex reality"

Challenges

  • Requires stable URIs
  • Use Linked Data for descriptive metadata or as a replacement of EAD/XML?
    • If former, becomes an additional layer to finding aid publishing
    • If latter, will need data remodeling

Steps to Prepare

  • Utilize @id to assign arbitrary ID's when possible
  • Use (inter)national authorities and utilize @source
  • Make your data as clean and consistent as possible

Future Research

Questions?

Resources

Finding Aid-LD: Implementing Linked Data in a Finding Aid Environment

By Jacob Shelby

Finding Aid-LD: Implementing Linked Data in a Finding Aid Environment

The slides to a talk I presented at Mashcat 2016.

  • 3,113