Data-Driven Insights into the Past

The Living with Machines project represents a groundbreaking multidisciplinary collaboration involving the British Library, the Alan Turing Institute, and universities including Cambridge, East Anglia, Exeter, Queen Mary, and King’s College London. Its mission is to explore historical narratives by employing data science, machine learning, and public engagement activities, including crowdsourcing tasks on the Zooniverse platform. A core focus has been analysing 19th-century newspapers to uncover insights into societal shifts during the industrial age.

In 2022 and 2023 the team ran a series of tasks in the citizen science platform Zooniverse. They asked members of the public to look at articles from 19th century newspapers that mentioned specific types of machine that they thought would yield interesting results.

Key Research Strands and Outputs for Public Engagement

The project includes two primary research strands, each with interactive notebooks designed to analyse data, visualise findings and engage diverse audiences:

The Language of Accidents notebook explores how mechanisation influenced accident trends. It offers visualisations of accident frequencies by year, geographic distribution, and victim demographics, such as age and gender. The tool also includes background information about the data sources, allowing users to contextualise their findings.

The Language of Mechanisation notebook investigates how words related to mechanisation evolved over time. It enables users to explore changes in word meanings by year, location, and newspaper publication, offering a rich view of language dynamics in the 19th century.

A Collaborative and Adaptive Design Process

The project team conducted detailed discussions to understand datasets, such as crowdsourced annotations and metadata from historical newspapers. A range of graphs were created in jupyter notebooks to beter understand the data and better understanding the direction to take.

Using a Miro board, the team mapped out iterative design solutions, ensuring alignment with project goals and audience needs. Various technical solutions were explored, with constraints like time and budget guided the decision to adopt existing notebook platforms, such as Observable and Jupyter. While Observable’s default design system introduced limitations, the team focused on enhancing user experience by refining the narrative structure, simplifying terminology, and testing chart types and interactions, designed to engage a broad audience. Accessibility considerations were prioritised, although some challenges, such as colour-based visual distinctions and browser-specific issues, were noted.

User-Centred Testing and Iterative Refinements

Usability testing was conducted in the latter stages of development through open-ended questionnaires, allowing stakeholders to clarify objectives, user needs, and to validate the notebooks’ design. Two rounds of testing, each with 8–10 participants, provided quick feedback on functionality and usability. Open-ended questions guided the sessions, highlighting key issues, which were promptly addressed. Both notebooks allow users to drill down to the original source documents, linking individual data points with contextual references.

Iterative adjustments continued to be incorporated based on feedback, ensuring the notebooks remained intuitive and engaging. This continuous refinement reflected the project’s commitment to delivering tools that met diverse audience needs, from quick overviews to detailed data exploration.

Conclusion

The Living with Machines project exemplifies how interdisciplinary collaboration and thoughtful design can reimagine historical research. By blending data science, public engagement, and accessible tools, it has illuminated new perspectives on 19th-century societal changes. The project’s success underscores the importance of integrating research, design, and user input to create impactful, lasting contributions to our understanding of history.