David Caldwell of The New England Wireless & Steam Museum kindly took the time to answer questions from Sara Brumfield of FromThePage, and discussed their project and experience using the platform.
First, tell us about your documents.
The New England Wireless & Steam Museum is a museum of engineering and technology with a collection of working steam engines, a larger collection representing the history of radio and wireless communication, and several historic buildings that include a 1907 wireless station that is on the National Register. We also have an uncatalogued library of more than 15,000 volumes from the 19th and early 20th centuries, roughly 75 percent of them on shelving.
To get a quick start on cataloging, we thought we could build a shelf list by transcribing the authors, titles and publishers from photographs of the shelves. So, our 'documents' are scans of book spines, 30 or so at a time, as they sit in our library.
What are your goals for the project?
First, to test our idea: Is this really practical?
Is it faster than working with the physical books? What fraction of the books won't be identifiable? Are there any benefits we hadn't expected?
Second, to get some real work done.
We chose one collection as our test case and set out to develop a first draft of a shelf list for 3,000 steam history and engineering books. This is our first step toward eventually having an online catalog.
How are you recruiting or finding volunteers/collaborators? Who is doing the transcribing?
We are an all-volunteer museum, so we recruited from our regular volunteers, people they knew, and others who have expressed interest in the library in the past.
Can you share your experience using FromThePage?
It was great.
We wanted to capture structured data and be able to download it when done, which took some experimentation.
The staff at FromThePage were very helpful. We took the time to write up thorough transcription guidelines. Once we opened the project up to our volunteers, it was straightforward for them to enroll themselves and get going with little guidance. Having an image viewer that let them zoom and rotate the images was important since the text on book spines can run in any direction.
We had captured shelf images for two-thirds of our test collection just as the museum shutdown for the coronavirus in the spring, and were able to get someone back into the building once later to photograph the rest. Because FromThePage is web-based, volunteers were able to work on the project from wherever they were holed up, and at a time when they couldn't have worked in the physical library at all.
In this test collection, we were able to identify almost 80 percent of the items, and we now know where each one is. We also know exactly where each of the 700 unidentified items are, so we can go directly to them on the shelf to complete the list when we get back into the library in person.
How does FromThePage & crowdsourcing fit into a museum context?
Well, we have 12,000 more volumes!
We're also considering ways to use crowdsourced transcription for cataloging hundreds of radios, most of which are in very dense storage, by taking photos of the front panels and builder's plates.
It is hard work to unpack the radios to take photos, in less than ideal conditions. But once that is done, other volunteers could transcribe the information from the photos. They could work anytime from anywhere, without having to visit the storage areas. By holding the images, keeping them associated with the transcriptions, and providing for editorial oversight, FromThePage makes organizing the work and managing the project much simpler than it would be with any scheme we could cobble together ourselves.
What would you tell folks considering a similar project?
Start small and try it. Take good clear photos.