Skip to content

Instantly share code, notes, and snippets.

@almibe

almibe/code4lib2021-slides.md

Last active March 24, 2021 17:13
Show Gist options
  • Star 2 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save almibe/c017fb8f60d9dc136832ff5af265bb0d to your computer and use it in GitHub Desktop.
Save almibe/c017fb8f60d9dc136832ff5af265bb0d to your computer and use it in GitHub Desktop.
Download ZIP
Raw
code4lib2021-slides.md
marp
true

Everyday Linked Data

Alex Michael Berry github.com/almibe twitter.com/alexmiberry

Goals

  • Explain how RDF and Linked Data are more general purpose than a lot of people give them credit for
  • Suggest adopting RDF for more common day to day data tasks rather than only large projects
  • Propose some ideas for moving forward with this and go over some issues I've ran into

Non-Goals

  • Teach RDF in 20 minutes
  • Make specific recomendations for software
  • Be 100% correct about everything I say :P

But First

  • Breifly explain what RDF and Linked Data are
  • Explain their role in libraries and archives

RDF (Resource Description Framework) in Two Slides

  • RDF was developed in the late 90's to help describe web sites, but its use has expanded since then (see https://bioportal.bioontology.org/)
  • RDF builds on a very simple subject predicate object model that is also very flexible

RDF in Two Slides

  • RDF is very modular, and a number of specs have sprung up around it to support:
    • Access/Modification/Query (SPARQL + HTTP specs)
    • Serialization (too many to mention)
    • Schema/Validation/Constraint (RDFS & SHACL)
    • Ontology Support/Inferencing (OWL)

Linked Data in Two Slides

from https://www.w3.org/2011/gld/wiki/5_Star_Linked_Data

Tim Berners-Lee, the inventor of the Web and initiator of the Linked Data project, suggested a 5 star deployment scheme for Linked Data. The 5 Star Linked Data system is cumulative. Each additional star presumes the data meets the criteria of the previous step(s).

Linked Data in Two Slides

☆ Data is available on the Web, in whatever format. ☆☆ Available as machine-readable structured data, (i.e., not a scanned image). ☆☆☆ Available in a non-proprietary format, (i.e, CSV, not Microsoft Excel). ☆☆☆☆ Published using open standards from the W3C (RDF and SPARQL). ☆☆☆☆☆ All of the above and links to other Linked Open Data.

RDF in Libraries and Archives

  • Usage of RDF is increasing in the libraries and archives world
  • Standards such as MODS/MADS have mappings to RDF
  • BIBFRAME is the elephant in the room
  • The Fedora repository and others implement the Linked Data Platform spec

RDF in Libraries and Archives

  • So there is a professional development aspect to learning about Linked Data and RDF

* That isn't what this talk is about

What is this talk about?

  • Using Linked Data/RDF for day to day tasks
  • Namely working with data that won't leave your computer/private server
    • Can still take advantage of existing Linked Data
  • Is this antithetical to linked (open) data?
    • I'd argue that it isn't since it will increase the usage of Linked Data

Examples

  • Work Stuff
    • Personal note taking
    • TODO tracking
    • Internal modeling
    • Analytics

Examples

  • Non-Work Stuff
    • Personal budgeting
    • Recipe management
    • Personal/Pet medical records
    • TTRPG Note taking

Examples

  • So kind of anything
    • Probably not best for working directly with large amounts of text or binary data though

Aspects of RDF Adoption

  • Education/Documentation
  • Tooling
  • Certain concepts can be overly challenging or esoteric

Education/Documentation

  • Well written specs
  • But there's a lot of outdated information
    • It feels like 1/2 of all links are broken when researching
  • Books are often outdated or very focused on a single aspect of Linked Data/RDF
  • Vendor specific/directly from the vendor (aka marketing)
  • Much is extremely ∀cadƎmic

Tooling

  • All popular (and most niche) languages have some support for RDF
    • I first learned about RDF via Prolog
  • Many stores/servers exist
  • It can be very hard to determine what exactly is supported by any given library/tool
  • A lot is propritary or inactive

Challenging/Esoteric Concepts

  • What are the best practices for crafting IRIs when you aren't talking about webpages?
  • Blank Nodes
  • Reification
  • Schemas vs Ontologies

Comparing RDF

  • RDF has strengths and weaknesses compared to the common gotos
  • RDF can be file based or server based
  • RDF has a common query language
  • RDF is graph based so cardinality, nested data, and sparse data are well supported
  • Linking different data sources is baked in

Comparing RDF

  • Has a simple enough data model that ORMs aren't needed
    • I think it's important when working with RDF programatically that you deal with the actual RDF data model and don't abstract it
  • Allows for a gradual adoption of a schema
    • Allows you to just start storing data and evolve a schema as you go if needed
    • But schema/validation/constraint support in tooling is currently lacking IMHO, I think tools need to become more opinionated in this regard

Advice for Adopting Linked Data

  • Have a specific project and goals in mind before looking into adopting Linked Data
    • Is validation important?
    • Do you need transaction support?
    • Do you need full text search?
    • What persistence do you need (Distributed DB/local files/memory)
    • Single or multi process/user access

Advice for Adopting Linked Data

  • Data modeling like software development is iterative, so be ready to experiment and make changes as you go
    • This includes both data modeling decisions as well as software choices
    • Look to existing standards and implementations for ideas, but realize your use case may require something different
  • Have fun

Thanks

Alex Michael Berry github.com/almibe twitter.com/alexmiberry

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment