National Library of Wales Crowdsourcing Platform

The National Library of Wales have a successful history of running crowd sourcing projects on their collections, and have also been IIIF implementers for several years.

In early 2017 they received funding from the Welsh Government to commission a crowdsourcing platform which would be:

• open source
• based around IIIF content
• support the W3C Web Annotation model (which was in the process of being ratified at the time)
• enable the creation of new crowdsourcing projects without costly setup time or technical expertise
• multingual for Welsh and English languages

The core requirement was for a sytem that was effectively a CMS (Content Management System) for crowd sourcing projects.

The system is hosted at: https://crowd.library.wales/en/s/home/page/welcome

The first project went live in January of 2018.

Projects to date:

• First World War Tribunals https://crowd.library.wales/en/s/war-tribunal-records/page/welcome
• Gwilym Livingston Evans (photographic collections) https://crowd.library.wales/en/s/livingstone/page/home
• Kyffin Diaries https://crowd.library.wales/en/s/kyffin/page/home
• DC Harries Portraits (photographic collections) https://crowd.library.wales/en/s/dcharries/page/home

The core of the platform has now been opensourced at:

https://github.com/digirati-co-uk/madoc-platform

With documentation at:

https://madoc.netlify.com/

Components

Capture Model:

JSON-LD document that describes:

• which fields to be provided
• what form the field takes:
  • dropdown
  • textbox
  • textarea
• which RDF vocab item the field maps to
• how the combination of fields are structured and presented to the end user

Annotation Studio:

https://annotation-studio.digirati.com/

• Can display IIIF content, based on Canvas Panel.
• Can convert capture model into a form for user input.
• Can serialise the captured information to a W3C web annotation
• Can GET and POST data to an annotation server.

Getting the content to other IIIF viewers

https://mcgrattan.org/oa_manifest?manifest=https://damsssl.llgc.org.uk/iiif/2.0/4696821/manifest.json&flatten=True&show_keys=True&most_recent=False&use_ids=False

Or direct to Mirador:

http://mirador.dlcs-ida.org/mirador?manifest=https://mcgrattan.org/oa_manifest?manifest=https://damsssl.llgc.org.uk/iiif/2.0/4696821/manifest.json&flatten=True&show_keys=True&most_recent=False&use_ids=False

Royal Society: Science in the Making

The Royal Society is the world's oldest scientific organisation, and scientific publisher.

They wanted to build a pilot project to expose their 350 year old archive, including:

• photographs
• manuscripts
• correspondence
• original illustrations and engravings
• referees reports

The system provides:

• generous interface
• user generated tags, transcriptions and comments
• navigation by theme
• search and advanced search
• fully IIIF compliant content

https://makingscience.royalsociety.org/s/rs/page/welcome

Photographic content: https://makingscience.royalsociety.org/s/rs/themes/fst01061714

N.B. the theme identifiers are Library of Congress LCSH/FAST identifiers.

Features:

• Timelines
• Map based navigation
• Thematic navigation
• Graph based navigation

Chronicle250.com

The Royal Academy was approaching the 250th anniversary of their first summer exhibition.

They wanted to build a site that would combine:

• Artworks
• Scholarly articles
• Indexes
• Catalogues

However, they had a problem when it came to providing these catalogues and indexes.

Over the 250 years of the exhibition, over 500,000 exhibits, and many hundreds of thousands of artists exhibiting. No way to manually catalogue or tag the artists in the catalogues.

We used:

• OCR
• Natural Language Processing (Named Entity Extraction)
• Linked Data
• IIIF Presentation and Image APIs
• Elasticsearch

to build:

https://chronicle250.com

A long form Medium article provides more technical information here:

https://medium.com/@matt.mcgrattan/the-royal-academy-summer-exhibition-chronicle250-com-dfc356a1ce8f