Code4Lib North: Day 1 Morning Notes

Morning presentations

Computational Thinking

  • Alan Harnum on Computation Thinking

Unfortunately, I was still dealing with the slight power bar issue, so I didn’t get a chance to blog most of it, but here are a few notes and the takeaways.

Coding does not teach you how to think computationally, so communicating with technology is more important in bridging and using a common language between the coders and non-coders.

Takeaways

  • Stop worrying about the imposter syndrome.
  • Technology is not computation
  • You probably already know this better than you think!

RDA for System Folks

  • Alison Hitchens

RDA is replacing AACR2 as the content guidelines for bibliographic & authority data to support resource discovery. Just telling you what data needs to be there. Also doesn’t need to happen in traditional catalogue.

Use of data to:

  • identify
  • select
  • obtain
  • clarify

RDA is not an encoding standard, it’s a display standard. Right now, encoding in MARC, but not necessary.

Implication for MARC

  • new fields and subfields to index and display esp. 264, separated information by 2nd indicator
  • new fields to replace the general material designation (GMD), what are you replacing it with? MARC fields 336, 337, 338 – effect on display? Suggest using icons
  • relationship designators – not new, but now highly encouraged in RDA
  • hybrid records: mostly because will not be retrofitting old records

Example: RDA in RDF at BNF

Some Issues

  • multiple pieces of information in one element?
  • need to use more identifiers
  • need to “dumb-down” library data?
  • RDA Toolkit is subscription based
  • need to better understand linked data

Bibframe

  • transition from MARC21 to more web-based, linked data friendly format
  • main classes: creative work, instance, authority, annotation

MARC, XML, JSON: And other stuff like that so we can all speak a common language and build the future together

  • MJ Suhonos

Slides

What’s the difference?

  • MARC, MARC XML, MARC HASH – serialization
  • MARC-8, ISO-8859-1, UTF-8 – encoding
  • AACR2, Dublin Core, RDA – schema

Every metadata format comprises of serialization, encoding, schema

  • ISO-2709 – can have multiple encodings
  • MARCXML  – can also have multiple encodings
  • MARC-HASH – passes, but may not support RDA

Ideal Metadata Format

Serialization

  • independent from schema and encoding
  • choose from multiple serializations from multiple serializations
  • XML-based formats are bad for this: PREMIS, MODS, EAD

Encoding

  • use UTF-8
  • otherwise part of serialization spec

Schemas

  • data-centric (like Dublic Core)
  • not markup centric (like AACR2, HTML)
  • test: if order or punctuation matters, not data-centric

RDF

  • Resource Description Framework for linked data
  • RDF = framework, not metadata format
  • combining schemas, encodings & serializations
  • multiple schemas in a single resource i.e. “record”
  • RDA Example:
    • serialization: RDF-XML
    • Encoding: UTF-8
    • Schema: RDA
  • DC-MODS Example
    • serialization: JSON-LD
    • encoding: UTF-8
    • schema: dcterms, modsrdf

Break Time

Pangolin Nap

Making (Web) Services Accessible

Slides

Listen to Your Library

  • Katie Legere

Even simple data, can be made more understandable with charts, and visualization.

Other ways: drawing information from sound. Gather data from listening.

Example: Etna Volcano activity turned into music

Allows us to discern differences quickly.

Took data from refstats (mangled version of libstats) and turned it into music by mapping library to instrument, volume to frequency, question type to melodic fragments (e.g. Beethoven Symphony No. 3). Used Finale to map each day to each bar.

The Revolution will be Preserved

  • Nick Ruest

Slides

Need to capture and preserve the website every day. Very easy to grab website using wget, but can get really large very quickly in WARC. Grab PDF screenshot instead using cron job.

Now what?

  • solution pack using Islandora. Drupal module to integrate with Islandora.
  • adds entries and derivatives, which are then indexed
  • WARC file which has screenshot, metadata, etc.

TODO

  • Solr indexing
  • Wayback Machine integration
  • automatic harvesting
  • automatic metadata harvesting

To Think About

  • where should something live, where metadata to come
  • we all have power, ethics, responsibility
  • Progressive Librarians Guild Toronto Area Chapter