Ran the same OCR models on 68 pages of historic newspaper. Every model hallucinated or looped.
DeepSeek-OCR-2, LightOnOCR-2, GLM-OCR β all melt down on dense newspaper columns.
You can try yourself using this @hf.co dataset: huggingface.co/datasets/dav...
23.02.2026 14:07
π 20
π 3
π¬ 4
π 0
Lovely videos from @bucksarchives.bsky.social showing value of archival work. ALSO a perfect demonstration of why older traditional NLP approaches to Named Entity Recognition (NER) fail in historical contexts where people and place are not always mutually exclusive, nor easily distinguishable.
10.11.2025 10:32
π 7
π 3
π¬ 0
π 0
Image of part of 1851 England Census return
Any ideas what the word/s between 'necessary' and 'Registrar' are in this marginal note from the 1851 English census? I thought I was pretty sure but keep doubting if I've got it right!
07.11.2025 09:12
π 0
π 0
π¬ 1
π 0
London Lives
Really pleased to announce the launch of the all-new, all-dancing, London Lives website - www.londonlives.org It has been thoroughly re-engineered to facilitate more types of search, and redesigned for phones and tablets. The team very much hopes peope like it. 1/
05.11.2025 11:24
π 146
π 114
π¬ 8
π 21
Was Manchester really as segregated as Engels said? What kept the rich and poor apart.... if anything? My first article is out today in @historicaljnl.bsky.social and I'm so pleased to share it with you all! doi:10.1017/S0018246X25101246
@stjohnscollege.bsky.social @camunicampop.bsky.social
21.10.2025 07:17
π 103
π 44
π¬ 9
π 5
High rates of polygyny do not lock large proportions of men out of the marriage market | PNAS
There is a widespread belief, in both the scholarly literature and the popular press,
that polygyny prevents large numbers of men from marrying by ...
π¨ The Economist has been telling you for years that polygamy causes civil war by locking men out of marriage. A new article with @rebeccasear.bsky.social and @anthrolog.bsky.social explains that the demography of marriage markets doesn't actually work that way. π§΅
www.pnas.org/doi/10.1073/...
06.10.2025 12:43
π 130
π 51
π¬ 3
π 9
A colour photograph of a hand holding a water damaged Nokia box. On it are three men in business attire playing football.
If you've ever wondered what life in an archive is like, we've discovered this 2005 mobile phone box we've somehow got that's a completely accurate depiction.
13.08.2025 15:23
π 22
π 1
π¬ 0
π 0
A BBC news headline which says βPenguin says it did βall necessary due diligenceβ with The Salt Path
Iβm no expert but I wouldβve got a person to do it.
09.07.2025 12:39
π 5638
π 1047
π¬ 150
π 84
Excited to try this out! π
07.07.2025 10:46
π 0
π 0
π¬ 0
π 0
A poster advertising the 'Wills Project Transcribathon' on Thursday 24 July, 1-4pm BST, in the Digital Humanities Lab at the University of Exeter, and on zoom. The poster features black text on green and yellow backgrounds, and three images - a box of folded will manuscripts, an unfolded will manuscript, and the painting Thomas Braithwaite of Ambleside making his will, Abbot Hall, 1607. Photo: Lakeland Arts.
Wow - so many people already signed up for our Wills Project Transcribathon later this month! π
If you drop by you can transcribe some lines & help make 25,000 English wills more accessible for all.
In person & online, join us here: willstranscribathon.eventbrite.com
#EarlyModern ποΈ #Palaeography
07.07.2025 07:09
π 43
π 32
π¬ 2
π 2
Promotional image advertising 'Wills Project Transcribathon, Thursday 24 July, 1-4pm BST, Digital Humanities Lab University of Exeter and on Zoom. The image also features a portrait of a man making his will, a photo of a box of wills and a page of a will.
π’WILLS TRANSCRIBATHON KLAXONπ’
Join us on Thursday 24 July to transcribe wills and go behind the scenes of the wills project!
In person and on Zoom - full details & registration here: willstranscribathon.eventbrite.com
#EarlyModern ποΈ #transcription @leverhulme.ac.uk @uniofexeterhass.bsky.social
24.06.2025 07:46
π 41
π 21
π¬ 2
π 0
π’ new paper klaxon π’
@felixkersting.bsky.social and I reconsider agrarian inequality in German and Swedish history and the role played (or not played) by landlordism in the road to fascism in Germany and democracy in Sweden. Iβm very happy about this one!
journals.sagepub.com/doi/10.1177/...
10.06.2025 20:11
π 25
π 4
π¬ 0
π 0
Welcome to the archives.
10.06.2025 08:02
π 2828
π 668
π¬ 95
π 171
Very proud to see @charmianmansell.bsky.social on this shortlist!
29.05.2025 10:37
π 2
π 0
π¬ 0
π 0
Thomas Turner β The Diary, 1754β1765
I have finished posting the 420,000 words in Thomas Turner's diary of 18thC rural life ( as transcribed by Dean K Worcester). This is significantly more than the published edition (Vaisey, 1984). www.thomasturner.org.uk It's a work in progress. #18c
12.05.2025 15:42
π 38
π 16
π¬ 4
π 1
Historians use data science to mine the past | PNAS
Historians use data science to mine the past
β¨β¨Maps & newspaper research from LwM & MRM teams (@danielwilson.bsky.social @joshrhodes.bsky.social @jonhistorian61.bsky.social @nottinauta.bsky.social @ruthahnert.bsky.social @npedrazzini.bsky.social @kallewesterling.bsky.social + more) featured in @pnas.org this week!
www.pnas.org/doi/10.1073/...
02.05.2025 12:51
π 20
π 11
π¬ 1
π 2
A map of part of the Slow Ways network, covering Wales and the middle of England. We can see all towns and cities connected by straight lines, revealing the structure of the walking network.
Just finished a day with the #SlowWays team in #Birmingham, charting big plans for a giant extension of our people-powered walking network.
Can't wait to share more.
β€οΈβπβ‘
07.04.2025 16:43
π 32
π 8
π¬ 1
π 0
Are there any UK based academics out there who've successfully claimed back withheld tax from a US fellowship? I'm trying to claim tax back from a Huntington Library fellowship and have no clue if I'm filling in the right forms! π€
04.04.2025 11:42
π 0
π 0
π¬ 0
π 0
Let me know if this makes sense! If not, happy to arrange a call.
25.03.2025 11:43
π 0
π 0
π¬ 1
π 0
4. Repeat this process for the osopenroads files (since some addresses are linked to only gb1900 and some only to osopenroads).
25.03.2025 11:43
π 0
π 0
π¬ 1
π 0
3. The scot_1851_gb1900_recidlkup.tsv file provides a bridge between the recids in the I-CeM file you've downloaded to the gb1900 map data in the scot_1851_gb1900.tsv. Once linked, you can map any of your variables in the downloaded I-CeM file using the geometry data in the scot_1851_gb1900.tsv file
25.03.2025 11:43
π 0
π 0
π¬ 1
π 0
2. Link the 'recid' field in your Govan I-CeM data to the 'recid' field in the scot_1851_gb1900_recidlkup.tsv file. Then link the 'gb1900_uid' in the scot_1851_gb1900_recidlkup.tsv file to the 'gb1900_uid' field in the scot_1851_gb1900.tsv file.
25.03.2025 11:43
π 0
π 0
π¬ 1
π 0
I-CeM
1. Download the Govan 1851 data from icem.ukdataservice.ac.uk
25.03.2025 11:43
π 0
π 0
π¬ 1
π 0
Hi Isaac, Great that AddressGB might be of use to you! I might be misunderstanding what you're saying but it sounds like you're trying to geo-code the AddressGB dataset, which is itself already geo-coded? If you're interested in mapping the 1851 census data for Govan, here's what you'd have to do 1/
25.03.2025 11:43
π 0
π 0
π¬ 1
π 0
Laughter as a Political Coping Mechanism schedule
Laughter as a Political Coping Mechanism schedule
Laughter as a Political Coping Mechanism schedule
Laughter as a Political Coping Mechanism schedule
Here's the schedule for 'Laughter as a Political Coping Mechanism', a conference that
@drhelenroche.bsky.social and I are organizing on March 28-29 for @durhamhistory.bsky.social. We'll spark fresh conversations about resistance to powerβand have a few laughs too! Please join us!
20.03.2025 22:50
π 7
π 4
π¬ 1
π 2
Full and Final Edition Now Available
News article - 24 February 2025
Big news! The digital edition of Alice Thornton's four Books is now complete and online. There are 1,019 pages (c. 270,000 words) of Thorntonβs life-writings in both modernised and semi-diplomatic versions and they are fully searchable.
#EarlyModern ποΈ π π
thornton.kdl.kcl.ac.uk/posts/news/2...
07.03.2025 09:22
π 67
π 37
π¬ 1
π 5