Living with Machines in the LibCrowds newsletter

We’ve been running crowdsourcing projects at the British Library for over a decade, and began using MailChimp to send a newsletter to subscribers in 2015, when we first launched the LibCrowds platform.

I’ve included updates on Living with Machines work with crowdsourcing in this newsletter since December 2019. As we start to write up our crowdsourcing projects, I thought it’d be useful to compile mentions of the project.

April 2023: Your help needed: How did machines change accidents in 19th century Britain?

Thank you to the thousands of volunteers who’ve helped us find news stories about accidents involving machinery, then recorded where those accidents happened, the age and gender of those affected, and the type of site where they happened, from fields to factories to home.

As we approach the final few months of the Living with Machines project we’re starting to work intensively with these records. Can you help us finish the tasks so that we can work with the full set of newspaper articles about accidents? 

They’re a great chance to see history in the making, and in n the words of Luke Hare (Research Data Scientist) and Mariona Coll Ardanuy (Computational Linguistics Senior Research Associate):

‘We will use your inputs, combined with cutting edge artificial intelligence techniques, to identify the locations where accidents occurred and analyse how this affected the public reaction.’

Find out more at https://livingwithmachines.ac.uk/how-did-machines-change-accidents-in-victorian-britain/

Join over five thousand volunteers at https://www.zooniverse.org/projects/bldigital/living-with-machines

Image: some of the tags that volunteers have added to 19th century newspaper articles on Zooniverse: the most prominent words being FatalAccident, railway, accident, rail, grisly, children, VictorianInventions, BoilerExplosion

Your comments needed: how would you use data from Living with Machines’ Zooniverse projects?

We’re busy finishing and documenting and sharing work from the Living with Machines project before it comes to an end this July.

We shared some data from our early Zooniverse projects on the British Library’s Research Repository back in 2020 (read the blog post ‘First crowdsourced datasets available’ or view the datasets https://doi.org/10.23636/1197).

You can get a sense of the range of past projects we’ve undertaken via blog posts on our website at https://livingwithmachines.ac.uk/?s=zooniverse.

Broadly speaking, we’ve looked at articles in 19th century newspapers for :

• collecting accidents involving machinery for linguistic and historical analysis, with further detail on the age and gender of victims mentioned, the geographic place and type of site (factory, mine, etc) where it occurred
• investigating exactly what kinds of things were called ‘machines’ via ‘describe’ (transcription) and ‘classify’ tasks
• sorting adverts from new stories about machines in ‘ad or not’
• understanding how specific words related to machines (or more precisely, vehicles) changed over time and place: car, trolley, coach, cycles

As we prepare to share data from our newer Zooniverse projects, we’re curious – how would you use Zooniverse data? Your comments will help us understand how to tweak what we provide in datasets (bearing in mind they’re all based on Zooniverse’s internal data structures). You can email us at digitalresearch@bl.uk with your thoughts, or comment on the Zooniverse forum.

December 2021: Your help needed: Living with Machines

Mia Ridge writes: Living with Machines is a collaboration between the British Library and the Alan Turing Institute with partner universities. Help us understand the ‘machine age’ through the eyes of ordinary people who lived through it. Our refreshed task builds on our previous work, and includes fresh newspaper titles, such as the Cotton Factory Times.

What did the Victorians think a ‘machine’ was – and did it matter where you lived, or if you were a worker or a factory owner? Help find out: https://www.zooniverse.org/projects/bldigital/living-with-machines

Your contributions will not only help researchers – they’ll also go on display in our exhibition!

Project update: Living with Machines Ad or not?’ task

Mia Ridge writes: In August we launched a new ‘ad or not‘ task. Previous crowdsourcing tasks had given us an incredible dataset to work with, but our analysis hit a snag. Advertisements were often repeated in successive issues of a paper, so machines that were heavily advertised were over-represented in our data. You can see the impact of ads in these initial visualisations – the sewing machine might be familiar to people who took part in our first ‘What was a machine?’ task in late 2020.

The technology that aims to identify them does not work very well on nineteenth century newspapers, so we asked people to help identify ads. This was the first time we’d designed a task to work in the Zooniverse app, and the results were astonishing, as each datasets was completed almost overnight. We’re now looking to use the results to train machine learning models to predict whether other text snippets are ads or not, then test those predictions with a Zooniverse task – this is an important and exciting next step for us. Overall, we saw nearly 30,000 classifications on our tasks this year. 

December 2020: Crowdsourcing officially launched – and completed! – for Living with Machines

Building on the lessons learnt from earlier experiments, in early December we launched two new crowdsourcing projects with data scientists from the Living with Machines project. These projects aimed to integrate linguistic research questions with tasks that encouraged volunteers to engage with social and technological history in the pages of 19th century newspapers. We tweaked the project after feedback from Zooniverse volunteers, and were delighted to be recognised as an official Zooniverse project.

Thanks to the mighty power of Zooniverse volunteers, the tasks were completed within a few days. Analysing the results will keep us busy in the first few months of 2021 and we look forward to sharing an update in our next newsletter.

May 2020: Wow! Thanks to all our contributors

At the time of writing, the number of all-time contributions on the LibCrowds platform has reached over 176,000 contributions from over 2400 registered volunteers. You’ve also helped us reach a new milestone, with tasks on over 100 volumes of playbills completed for In the Spotlight. Our two tasks on the Zooniverse platform for the Living with Machines project were completed in early March and the data is being prepared for analysis.

Thank you for all your help! If you’ve done even one task on one of our projects lately, you’re part of a truly international movement that stretches from Gaborone to Auckland to Moscow to London. We can’t thank you enough for the time and effort you’ve put into helping researchers and making historical collections more accessible.

December 2019: Introducing… Living with Machines