Connecting people to places using library data – Tim Sherratt

Many people come to libraries with questions. Sometimes they are academic and sometimes they are personal. What can you tell me about my house? What did my suburb look like before the townhouses took over, before the freeway cut through the farmland? At State Library Victoria, staff hear questions like these nearly every day. The answers are stacked on shelves and stored in online repositories: maps and directories, planning documents, photographs of streets now changed beyond recognition, newspapers that tracked the history of regions.

In practice, answering questions about a place can mean hopping between catalogues, scrolling through directories one page at a time, cross-checking old street names and comparing landmarks on hundred-year-old maps with those on modern geographic tools. The information is abundant, but it resists being gathered into a single view. The more immediate answer to ‘what used to be near this spot?’ is missing.

It was this gap, between what the Library held and what people wanted to know, that framed Tim Sherratt’s 1 residency with SLV LAB.

The collection material

Tim arrived at State Library Victoria in late September 2025 as SLV LAB’s second Creative Technologist-in-Residence. The lab’s residency program welcomes practitioners like Tim to work with staff to explore new ways of connecting people with the Library’s collection – vital work to meet the ever-evolving needs of library users. He had 12 weeks to prototype some solutions to the question ‘what can you tell me about this place?’.

This challenge was posed to Tim weeks before he officially began his residency, at Wikifest 2025, a workshop that gathered technologists, researchers and cultural workers to explore the potential linkages between Wikimedia and GLAM 2 collection data. A librarian in Tim’s group described the familiar interaction with visitors trying to piece together the history of a place. Having spent decades listening to the questions people ask of archives, he began to formulate some approaches.

State Library Victoria’s collection material – and its sheer volume – suited the challenge. For example, the digitised Sands & McDougall directories alone contain around seven million entries 3, recording commercial, industrial and residential land use across Victoria from 1860 to 1974. The Committee for Urban Action collection holds tens of thousands of photographs taken in the 1970s when fears about the loss of architectural heritage prompted volunteers to document streets before they disappeared. Mahlstedt fire insurance plans and MMBW (Melbourne and Metropolitan Board of Works) water and sewage plans reveal the features of streets and buildings of the time in great detail; parish maps chart the administrative boundaries of the colony and the state.

These collections were created to describe where things are, yet much of that spatial information has been flattened in the process of digitisation: maps are images, directories pages are scans, photographs are described in text. Tim’s task was to coax the latent geography back out.

About Tim Sherratt

Calling Tim Sherratt a historian undersells him slightly. Calling him a hacker does the same. Over the past thirty years, he has moved between research, cultural institutions and code, building tools that sit in the space between official infrastructure and individual curiosity. In 1993, in the early days of the web, he worked on the Australian Science Archives Project based at the University of Melbourne, which led to the development of Australia’s first archives website. This sparked his interest in GLAM collections and how the affordances of the then fledgling web could allow people to access history and communicate about it in different ways.

In 2007, while at the National Archives of Australia, Tim realised that applying computational methods to collections could open new avenues for analysis and exploration – an insight that became a major turning point in his career. Since then, his projects have ranged from dashboards (PROV Data, NAA Digitisation) and searchable digitised journals, to political interventions (The Real Face of White Australia, Historic Hansard) and playful data experiments (Operation Random Words, redacted). Most recently, he was Associate Professor of Digital Heritage at the University of Canberra, and before that, Manager of National Library of Australia’s online research portal Trove. 4

At the centre of Tim’s practice nowadays is the GLAM Workbench, a sprawling, carefully documented collection of scripts, notebooks and examples designed to help people work with data from galleries, libraries, archives and museums. The Workbench reflects his philosophy: that institutions cannot do everything themselves, and that access depends on a commitment to openness.

Starting out with parish maps

Early in the SLV LAB residency, Tim spent time talking with librarians and collections staff who guided him through the Library’s holdings. They pointed him towards collections with untapped potential, explained how and why certain metadata had been added and flagged known frustrations.

One of these conversations was with Suzie Geermans 5 (Senior Librarian, Acquisitions and Description), who showed Tim the Library’s parish maps and explained that around 66% of these catalogue records included coordinates for each parish. The Library’s cataloguers had added the latitude and longitude of the parish centre by looking it up in a government gazetteer 6.

Tim realised that if you combined those coordinates with two other pieces of information already in the record – the physical size of the map and its scale – you could calculate the area each map covered. A bit of trigonometry later, he had an approximate bounding box for every map.

When he plotted the results, something magical happened. The individual rectangles assembled themselves into the outline of Victoria – because of years of careful cataloguing. This moment captures Tim’s process in a nutshell: he looks for what is already there, waiting to be connected.

Learn Tim's method for drawing bounding boxes for parish maps in the Library's collection.

When data refuses to behave – The Committee for Urban Action photos

Not all collections cooperate so well. While the Committee for Urban Action photographs are rich in detail, they weren’t created with digital mapping in mind. Each photograph is titled with a street name, suburb and sometimes an orientation, but there are no coordinates.

Tim’s solution was part engineering, part detective work. He turned to OpenStreetMap, a community-maintained map of the world that includes detailed information about streets and intersections. By matching the street names in the photo titles with OpenStreetMap data, he could identify likely intersections and road segments. He then attached that spatial information to the photographs.

It worked surprisingly well. Most of the images could be placed on a map with reasonable accuracy, although some resisted. Street names had changed, and suburb boundaries had shifted, and there were some minor inconsistencies in the photo titles (but considering that the photos were taken and labelled by several volunteers across various suburbs and towns across the state, the labelling is, overall, remarkably consistent). Tim’s goal wasn’t perfection but usability; he accepted any imperfections as part of working with historical material. After all, a map that lets you explore thousands of photographs spatially, even with gaps, is a powerful thing.

Try out Tim's prototype CUA Browser

Making directories behave like data

The Sands & McDougall directories presented a different kind of challenge. Spanning more than a century, these volumes record who lived and worked across Victoria. They are invaluable to genealogists, historians and anyone researching property history. They are also cumbersome to use online.

Although twenty-four volumes have been digitised 7, they were originally accessible only as scanned pages or PDFs. Searching across volumes was impossible. Finding a specific entry meant scrolling page by page hoping not to miss it.

Tim’s approach was to break the pages down to their smallest meaningful units. Using OCR (Optical Character Recognition) 8 data embedded in ALTO 9 files, he indexed each line of text. This allowed searches to jump directly to an individual entry rather than a whole page.

As a result, you can now search for a name or address, and the interface takes you straight to it, highlighting the entry in context. These directories now behave more like a database and less like scanned books.

Learn how Tim created a new searchable interface for the Sands & McDougall directories

Learn how Tim used IIIF and ALTO to create thumbnail snippets for Sands & McDougall entries on a new interface

Letting maps speak to the present

One of the most publicly visible outcomes of the residency was Tim’s work on georeferencing maps on Allmaps, a browser-based tool that links historical maps to modern geography.

Georeferencing 10 allows you to overlay an old map onto a contemporary one, aligning landmarks like rivers, roads and boundaries so you can see how a place has changed.

Tim set up a system that allowed anyone to georeference maps from the State Library’s collection themselves. It was an open invitation to participate. By the end of the residency, over 900 maps had been georeferenced through this crowdsourced effort. Some volunteers focused on Victoria, others wandered further afield, mapping European battlefields or distant towns. The geography of interest expanded as soon as the tools allowed it.

Learn how Tim turned the Library’s maps into data with Allmaps and some ‘data plumbing’

Learn how to georeference maps in State Library Victoria's collection

Charting the rise and fall of Victorian newspapers

In the final days of his residency, Tim turned his attention to something less obviously visual but just as deeply tied to place: newspapers. Victoria’s publishing history is rich and scattered, with many local and regional titles sitting outside Trove, some digitised, others existing only as records of their publication.

Tim built a database that linked titles to the places they served and the periods they were active, then plotted them on a map. The result is a way of seeing journalism geographically, as something that clustered around towns, and appeared and vanished as communities grew or declined – a boon for local historians.

Learn how Tim developed a new app to explore Victorian newspapers

Try out the newspaper explorer

Making My Place

By the end of the twelve weeks, Tim had developed a compelling suite of prototypes, many of which could be accessed on a web interface he developed called ‘My Place’. It lets users enter an address or place name and see related directory entries, maps, photographs and newspapers drawn from the library’s collections.

Try out My Place

Tim is very careful to point out that these apps aren’t finished products. Rather, they are a way of thinking out loud. They demonstrate what is possible, invite feedback and encourage conversations about next steps.

The benefits of Tim’s work are clear. Librarians can answer questions more efficiently, researchers can move faster and members of the public can explore their own histories without needing to understand how collections are organised behind the scenes. Tim’s work also suggests a shift in how cultural institutions might think about access. Search, in this model, is not limited to words, but includes proximity and overlap. It allows us to ask, implicitly or explicitly, what was happening around here and when.

For State Library Victoria, the residency has opened conversations about metadata practices, about the value of spatial description and about the kinds of questions users bring to the Library.

"Tim’s work will have immediate benefits for supporting researchers, and his experiments demonstrate the opportunities for how we can find new ways to provide access to Library collections.” 

– Tom Vasey, Senior Librarian, Digital Access

There are also implications beyond the Library. Urban planners, genealogists, environmental researchers and artists all work with place-based questions. Tools that allow historical data to be layered, compared and navigated spatially invite cross-disciplinary use.

Underlying all of this is Tim’s insistence on openness. The code, the documentation, the datasets are available for others to adapt. The work is designed to be taken apart and reused – this is especially useful in a sector constrained by bespoke systems and limited resources.

Why bother?

In a talk to staff delivered at the end of the residency, Tim posed a question that doubles as a manifesto: why bother? Why spend time extracting bounding boxes from catalogue records or wrestling with inconsistent street names? Why build tools that may never be formally adopted?

The answer for him lies in the people who come to libraries with questions and in the belief that collections should meet them where they are. It also lies in the satisfying moments when disparate data snaps into alignment, when the past reveals itself and becomes navigable. Tim describes himself, with self-deprecation, as an ageing hacker. But the SLV LAB residency demonstrates the enduring value of that identity. Hacking, in this sense, is not about breaking things, but about making them more useable and humane.

As the Library’s digital team begins the work of turning prototypes into tools, the influence of the residency will persist as a way of thinking. Place matters. Metadata matters. And sometimes the most transformative changes come not from building something new, but from learning how to see what you already have.

The full repository of code, tutorials and datasets can be found on the ‘My Place’ Experiment page on the SLV LAB website. You can follow and support Tim’s work at timsherratt.au. This SLV LAB residency was funded by the Maps that made Melbourne campaign.