Keynote 'Evolutionary Innovations: Collections as Data in the AI era' for Making Meaning 2024

My slides for #SLQMakingMeaning #CollectionsAsData, 'Evolutionary Innovations: Collections as Data in the AI era', are online at https://zenodo.org/records/10795641

‘Collections as data’ describes the movement to publish open data from museum, library and archive collections that began in the noughties. The benefits of machine learning for better discoverability and research with digitised/born digital collections are alluring. And the popularity of generative AI – and an increased awareness of the biases it reinscribes – has focused attention on responsible computational access to collections – but what does this mean in practical terms? Mia will share examples from the British Library and the Living with Machines data science project.

2022: an overview(ish)

A work-in-progress post about what I got up to last year.

The biggest thing I did in 2022 was co-curate an exhibition at Leeds City Museum for the British Library and Living with Machines project.

November: I was invited to the Archives nationales de France conference 'Crowdsourcing et patrimoine culturel écrit', where I spoke on Crowdsourcing as connection: a constant star over a sea of change / Établir des connexions : un invariant des projets de crowdsourcing par Mia Ridge, British Library, Royaume-Uni

In December I gave an online keynote on 'Citizen Science as Public History?' for the conference 'When publics co-produce history in museums: skills, methodologies and impact of participation' at The Luxembourg Centre for Contemporary and Digital History (C²DH), University of Luxembourg.

Crowdsourcing workshop activities: ideation and elaboration

I've been working on structures for online workshops for people working on crowdsourcing and other digital participation projects for museums, libraries and archives for over a decade now, learning from each institution I work with. I thought I'd share one of the slide decks I'm currently using.

The deck is labelled 'Coming up with and developing crowdsourcing ideas'. In a workshop or class on crowdsourcing it usually comes after sessions that explain the whats and whys of crowdsourcing in cultural heritage. It's designed to get people quickly working on practical ideas, anticipating issues and ensuring that their projects will fit into their specific institutional context.

The prompts currently include: What does success look like? Which audiences are interested? Why? What could you learn from trying this? Which collections are involved? Links to mission? Pros? Cons? How could you ensure data quality? Costs  (staff, tech)? Dependencies / assumptions? What problem does it address? Questions, concerns? What volunteer skills, experience needed? What will they learn? What tech, data is needed?

You can develop your own prompts based on the attributes that are important to you. The Collective Wisdom Handbook is a useful guide to figuring out what's important to you, from data quality to integration with existing workflows.

I mostly recently used this for a Europeana-funded workshop for the Estonian War Museum – General Laidoner Museum in March 2022.

The museums have shared some lessons from the workshop in a post for Europeana. Their report 'Estonian museums' experience in the field of crowdsourcing' not only provides some background on volunteering and crowdsourcing in Estonian museums, it also shows how they applied the prompts.

Crowdsourcing workshop activities: ideation and elaboration by Mia Ridge is licensed under CC BY-SA 4.0

Living with Machines exhibition launched

For the past year I've been co-curating the Living with Machines exhibition with John McGoldrick and working intensively with many others at Leeds Museums and Galleries and the British Library. It's inspired by the Living with Machines research project, and very much shaped by my interactions with volunteers on our Zooniverse crowdsourcing projects.

I've written a blog post These are a few of my favourite things… in the Living with Machines exhibition for the Living with Machines blog that explains something of the challenge.

I've also deposited our interpretation for wall panels and object labels for 'Living with Machines: human stories from the industrial age' at the British Library research repository.

The Living with Machines exhibition was open until January 2023 at Leeds City Museum.

Spot me and an exhibition object on the front page of the weekend Yorkshire Post
LwM team absorbed in the loom demonstration - Living with Machines exhibition opening
My Flickr album of installation and opening night photos

Official photography from Leeds Museums and Galleries

Me and co-curator John McGoldrick

Hidden REF nomination for crowdsourcing at the British Library

In September 2021 year I was delighted to be nominated for a Hidden REF award. The Hidden REF is a project that celebrates the work of people who are vital to the success of research, but who may go unrecognised by traditional academic criteria for research outputs. 

I'm sharing a copy of the nomination for LibCrowds, the platform and community on which In the Spotlight, a project crowdsourcing the transcription of historical playbills, was built:

LibCrowds is a platform dedicated to hosting crowdsourcing projects aimed at enhancing access to British Library collections. Since launching in 2015, it has hosted 171 projects, drawing in 265,000 contributions from nearly 3,000 registered volunteers, and many more anonymous individuals. The crowdsourcing projects greatly enhance the discoverability of library collections.

Our community of volunteers have contributed to projects such as: Georeferencer‚ providing more accurate, diverse metadata about digitised historic maps; In the Spotlight‚ transcribing 18th-19th century playbills (making them more findable and searchable); Convert-a-Card‚ retro-converting printed card catalogues into electronic records, particularly improving access to Chinese and Indonesian collections.

The platform is carefully designed for productivity; it's easy to use and interact with images. However, engagement with collections is also a key outcome. LibCrowds has built a strong community. Our surveys indicate that most contributors participate because it's enjoyable, and some take a personal interest in the subject matter. They can discuss discoveries with others through a forum, and can easily share images via social media.

LibCrowds has enabled important research findings. For instance, the playbills project has allowed research on plays which were previously important but which waned in popularity, and has revealed details about marginalised groups including women and Black actors. We are aware of multiple doctoral students working on aspects of theatrical history and researchers in several universities that have used the transcribed collections in their publications.

The scholarly and professional literature recognises LibCrowds to be an extremely valuable case study of a successful crowdsourcing project. It's referenced in dozens of articles and conference papers. Recently, insights from LibCrowds have been integral to the planning of research in the Library and Turing Institute's Living with Machines project, using crowdsourcing to engage the public with data science methods and produce effective and timely results about 19th century newspapers.