Should a volunteer use #AI to help them transcribe pages for a #crowdsourcing project? That question got me thinking about why, exactly, my answer is "no" and what kinds of purposes different transcriptions may be used for. content.fromthepage.com/can-voluntee...

We did a serious analysis of our experiments adding AI coding agents (Codex, Github Copilot Agent) to our development process at FromThePage for a group of friends in software yesterday, which I thought I'd share here as well. After a couple of months of experiments, the results are very mixed.

A little feature we shipped yesterday in FromThePage: you can now reference a work by page ranges, so fromthepage.com/unclibraries... shows a single two-page letter from the fifty-page folder uploaded as a single work!

What's the Character Error Rate of a Volunteer? Analyzing accuracy in cultural heritage crowdsourcing projects | @benwbrum.bsky.social | #DH2025

Sara and I were interviewed for the SAA Archives in Context podcast! archivesincontext.archivists.org/2025/06/27/s...

Sitting in coffeeshop listening to three strangers at the bar getting into a detailed conversation about the different accents of Louisiana in both English and French. It's a good start to the day.

Come hear Dreanna Belden share about "One Great Document: a Woman’s Perspective on the Marquis de Lafayette’s 1824 Visit to Yorktown and Norfolk," during our lightning talks at next week's ADE Annual Meeting.

Does Oxford University Press no longer publicize conference discounts or holiday sales?

I'm really enjoying the inaugural Alliance for Texas History conference at Texas State University. Best session I've attended so far? It's a toss-up between one on a freedman blacksmith and another on highway construction through African-American neighborhoods in Houston roadstaken.org #ATxH

Our first transcribathon event is today!

Crowdsourced transcription platform FromThePage has long championed public participation in historical research. Find out how they used Transkribus' AI technology to help their volunteer transcribers in this Success Story 👇 blog.transkribus.org/en/fromthepa... @benwbrum.bsky.social

Like many other #NEH -funded projects. the #sferaproject received notice yesterday that our grant has been cancelled. It's not yet clear what that means for the project, but we hope to have more info soon. Meanwhile, here's a screenshot of what they cancelled (our website-in-development):

A total of 44,091 manuscript pages have now been completed by Transcribe Bentham volunteers, with a further 361 in progress, and 10 awaiting review 👏

Excited to learn that our proposal for the Texas Conference on Digital Libraries was accepted! I still can't believe how lucky I am to have an amazing digital library conference an hour's walk from my house. #TCDL2025

Has anyone experimented with automated hand or script identification using #HTR or multi-modal #LLM AI tools? It would be interesting to locate pages within a corpus written in multiple hands--especially if the scripts were different enough to indicate different social classes of writers.

ISAC (formerly OI-UChicago) is offering an eight-week course in #Hieroglyphic #Luwian online through their continuing education program, May 29-July 24. I took the Sumerian class they offered a couple of years ago and was impressed -- it's well worth the $400. www.eventbrite.com/e/introducti...

At some point in the next few years, I'd like to take a course on Old English. As a 50-year-old with a career, it'll probably need to be online. The "Ancient Language Institute" offers one, but I've never heard of them. Any advice? #classicists? #medievalists?

August 24(?), 1790 exerpt from the log of the sloop Dolphin, kept by Capt. Benjamin Paddock, Jr. on a whaling voyage from Nantucket to the Bahamas and down to the West Indies, 1790-91. fromthepage.com/nharl/ships-...

This call permits commercial subcontracting! Needless to say: reach out to me if you are in need for a tool builder experienced in all things #DigitalHumanities, #IIIF, image- and text annotation, web maps, and #UI/#UX for #DH 🙂

Hello all manuscript fans! As we are approaching St Brigit's day, why not transcribe some of Cogitosus' Vita S. Brigidae in her honour? Images from the Cambrai manuscript ready to go here -> fromthepage.com/nicolev/eih-... #Brigit #medievalsky @historytcd.bsky.social @researchireland.bsky.social

We are offering two FREE, online semi-intensive short courses in #Phoenician and Punic #language in April 🏛️ 🏺 Elementary course info: ics.sas.ac.uk/events/eleme... Intermediate course info: ics.sas.ac.uk/events/inter... Share and Apply by 22 March!

One bright spot in my life is that "What's the Character Error Rate of a Volunteer?", my short paper for #DH2025 was accepted. If nothing happens, I'll see a lot of you #crowdsourcing #digitalhumanities people in Lisbon.

GLAM folk, got perspectives on 'Participatory Science' (or citizen history, citizen science, crowdsourcing) to share for a special edition of Citizen Science: Theory and Practice (CSTP) Journal? Abstracts to forms.gle/tZmaSEPixgAU... - by February 28 More: participatorysciences.org/2024/11/07/c...

Is there a location for the 2025 #dlfforum?

Another option is to include a range, e.g. 1480-1490, but that suggests that the date can't be 1479 or 1491. It's a real problem that most Library databases aren't equipped to solve yet.

My DigiKult keynote has just gone live! 'Hold steady: centering the human in the hype of AI' was my view of the potential and pitfalls for AI in libraries, museums and archives as of early October 2024 https://www.youtube.com/watch?v=voDWMB-903k&list=PLKrB8_QR_fyHm79rosJ8LDZOjiri07JXR&index=8 #AI

Using AV in class? AVAnnotate is hosting bi-weekly conversations about using AV in class. The first session (1.21 at 2pm CT) is working with AV archives: links, IIIF manifests, transcriptions, basic tags, structuring assignments. RSVP for the Zoom link: docs.google.com/forms/d/e/1F...

I just opened Blueskye and read this out to my wife. Where else could I find this? I love this site!

Historian and paleography friends! I have a hybrid Spanish-Nahuatl document I am trying to catalogue. It is brief. It look like possibly a property claim? If you have a moment to help me make it a title and approximate date, I'll thank you with a bevvie at our next conference.

ODH is participating in a new NEH/AHRC program for joint U.S.-U.K. projects undertaking humanities research on the ethical, legal, and societal implications of AI. For more information, please visit NEH’s International Opportunities page: www.neh.gov/initiatives/international-opportunities.

How would you interpret this envelope for a 1603 Spanish document? It looks like two different hands to me. Visible words include "Cartaxena" and "El governador don Gieronimo"

What happens when you combine the best open-source HTR model (Qwen 2 VL) with the best open-source zero-shot classification NER model (Knowledgator's fine-tune for Gliner)? Caracal. Try it now on Hugging Face : lnkd.in/e5bUZm-w #CulturalHeritage #DigitalHumanities #MachineLearning #HTR #NER

Today I'm launching #liiive: a new IIIF viewer that puts real-time collaboration front and center. View, annotate, and discuss together as if you're in the same room. Just paste a manifest URL and go! liiive.now 🎄 #DigitalHumanities #IIIF #DH #ArtHistory #Museums

How common were dual-use water mills in 19th-century Virginia? I'm marveling over this account of "Cotton & Logs sent to Doct Miller's mill". #skystorians fromthepage.com/benwbrum/jer...

Have you ever wondered where an author lifted a passage from when they didn't cite a source? @lucianli123.bsky.social was able to use LLM embeddings to trace sources Charles Darwin used in Origin of Species, as well as later authors quoting (or paraphrasing) Darwin! See his work at today's webinar:

What's the CER of a human? We are researching this currently, and I finally found the paper which inspired the work: Oliveira and frederickaplan.bsky.social's "Comparing human and machine performances in transcribing 18th century handwritten Venetian script" infoscience.epfl.ch/server/api/c...

I spent the last 5 hours on eldercare and family time, but managed to get a release out the door tonight: FromThePage 24.12-pre, a pre-release version of the codebase with an upgrade to Ruby 3 and Rails 7. github.com/benwbrum/fro... That ends #DayOfDH2024 for me.

making my FIRST EVER bsky post to show y'all photos from a trip i took over break during an MLA fellowship i'm on for the next year!! collected images and audio recordings--also visited Sul Ross State's Archive of the Big Bend--around the Permian Basin and Trans-Pecos

Testing a bug fix between meetings. This might be a good time to show off the slider feature we built into FromThePage's transcription screen. (It's nicer on a big monitor, but even on my laptop, adjusting the image view vs. transcription editor makes a difference.) #DayOfDH2024

This #DayOfDH2024 is on Monday, when we have our weekly planning meeting for FromThePage. We gather stats for our scorecard, so I've filled in the product side. (Interestingly, the number of pages transcribed last week fell by a third -- was it the US holiday or did a big project finish?)

Finished my overview of open pull requests for the team. Sadly, most of the open issues are blocked by me, and I'll be in meetings most of the #dayofdh2024

Starting my #DayOfDH2024 with a coffee and looking over the 21 open pull requests on the FromThePage repository to see what must be tested today before we cut a release.

I need help deciphering this passage in Jeremiah White Graves's 1852 financial accounts with "my Henry" -- Henry Washington, an enslaved adult laborer. For context, an earlier entry records "To cash lent you in Lynchg for Traces" "Trace chair"? "Trace chane"? ???

On Dec 12, join us for a webinar with Lucian Li, where he'll showcase a new method for identifying intellectual influences in historical texts using sentence embeddings. The webinar is at 12PM EST, 11AM CST, and 9AM PST. Sign up here: content.fromthepage.com/dec-2024-web...

Why do people transcribe @librarycongress collections? What do volunteers get out of crowdsourced transcription? Our article "Giving back, learning, relaxing, & having fun: personal motivations & impacts of a virtual volunteer transcription program" is out now! rdcu.be/dXXRs @tjowens.bsky.social

Have any quantitative studies been done on accuracy of crowdsourced manuscript transcription in practice--that is, when performed by volunteers engaged with the material, rather than offshore hires or mechanical turkers?

The new #NEHGrants are out! www.neh.gov/sites/defaul... We're pleased to be part of the SubjectSpotter team, who will be working to make emerging AI-based tools to recognize and identify entities within historic documents more accessible to librarians, archivists, and scholarly editors.

Looking for a job with impact? Join the British Library as a Digital Curator, OCR/HTR and help make the Library's collections more accessible to readers and computational methods! blogs.bl.uk/digital-scho... Closes Sunday 21 July 2024. Pay close attention to questions in the application process.

I'm reading a lot about #ChatGPT4o and #htr, so I figured I'd spot-check it with this document: fromthepage.com/lva/va-revco... The results were better than nothing, but still not great.

The UC Press sale includes one of the best books on the development of writing systems I've read in the last couple of years: Gordon Whittaker's Deciphering Aztec Hieroglyphics. (Don't worry, only the last section requires knowledge of Nahuatl) www.ucpress.edu/book/9780520...