Search the catalogue
100 results
- tool · 020
Mistral OCR AI
Commercial cloud OCR that converts scans into structured text in about 170 languages, cheaply per page for bulk jobs.
Paid · Some code 0 research · 2 corpus · 1 tested - tool · 071
World Historical Gazetteer
Free online index of historical place names worldwide. Look up places with their old names, match your own place lists against it, and publish them as linked open data.
Free · No code 0 research · 4 corpus · 0 tested - tool · 060
Gephi
Desktop program for drawing and measuring networks, such as who wrote to whom. The most common no-code network tool in digital history.
Free · No code 3 research · 0 corpus · 0 tested - tool · 050
Voyant Tools
Browser-based reading environment: paste in texts and get word frequencies, trends, keywords in context and visualisations. A common first step into text analysis.
Free · No code 3 research · 0 corpus · 0 tested - tool · 015
eScriptorium AI
Open-source web platform for transcribing historical manuscripts and print: segment pages, run or train models, correct and export. Strong on Hebrew, Arabic and other non-Latin scripts.
Free · No code 0 research · 2 corpus · 1 tested - dataset · 103
Trove newspapers
National Library of Australia portal to hundreds of millions of digitised Australian newspaper articles, mostly 1803–1954, with OCR partly corrected by volunteers.
- dataset · 105
Canadiana
Canada’s largest open collection of digitised historical print and archival material, including Héritage’s millions of pages of microfilmed Library and Archives Canada records.
- dataset · 102
Chronicling America
Library of Congress collection of millions of digitised US newspaper pages up to 1963, with freely downloadable OCR text. The default corpus for large-scale work on American newspapers.
- dataset · 106
HathiTrust
Shared digital library of more than 18 million books and serials from research libraries. Full-text search across everything; full reading of public-domain works.
- tool · 051
MALLET
Command-line toolkit best known among historians for topic modelling: finding groups of words that occur together across thousands of documents.
Free · Some code 3 research · 0 corpus · 0 tested - dataset · 101
Old Bailey Online
Searchable transcriptions of almost 200,000 trials at London’s central criminal court, 1674–1913, tagged for offences, verdicts and punishments, with page images.
- tool · 070
QGIS
Free desktop GIS. Historians use it to georeference scanned old maps, geocode places from censuses and directories, and trace historical features in layers.
Free · No code 1 research · 2 corpus · 0 tested - tool · 072
Recogito
Browser tool for annotating texts, images and maps together, especially tagging place names and linking them to gazetteers so they can be mapped. Recogito Studio is its newer, institution-hosted successor.
Free tier + paid · No code 1 research · 2 corpus · 1 tested - dataset · 104
SlaveVoyages
Voyage-by-voyage database of about 36,000 trans-Atlantic and 10,000 intra-American slave-trading voyages, built from archival records, with estimates of the trade’s full size.
- tool · 081
StoryMap JS
Free browser tool for slide-by-slide stories tied to points on a map. Its “gigapixel” mode lets you annotate one large image, such as a scanned historical map.
Free · No code 2 research · 1 corpus · 0 tested - tool · 013
Kraken AI
The open-source recognition engine behind eScriptorium, run from the command line. Built for historical and non-Latin material, including right-to-left scripts.
Free · Programming 0 research · 2 corpus · 2 tested - tool · 012
Tesseract
Free, long-standing program that turns images of printed pages into searchable text. Best on clean, high-resolution scans; old or damaged print usually needs clean-up first.
Free · Some code 0 research · 2 corpus · 6 tested - tool · 014
Transkribus AI
Upload scans of handwritten or printed documents and get transcriptions back, using public models or one you train on your own transcribed pages.
Free tier + paid · No code 0 research · 2 corpus · 2 tested - tool · 042
spaCy AI
Python toolkit that, among other things, finds names of people, places, organisations and dates in text.
Free · Programming 0 research · 1 corpus · 2 tested - tool · 027
Whisper AI
Speech-recognition model that transcribes oral history interviews and archival audio, with timestamps, in many languages.
Free tier + paid · Some code 0 research · 1 corpus · 2 tested - tool · 011
ABBYY FineReader PDF
Commercial desktop OCR that libraries have used for decades to digitise newspapers and books. Has a dedicated mode for Gothic and Fraktur type.
Paid · No code 0 research · 0 corpus · 3 tested - tool · 016
Chandra AI
Vision-language model that turns scans and PDFs into structured text, keeping tables and layout. Chandra 2 covers about 90 languages.
Free tier + paid · Some code 0 research · 0 corpus · 2 tested - tool · 018
DeepSeek-OCR AI
Small open model that compresses page images for fast bulk conversion. Built mainly as research into compression rather than careful transcription.
Free · Programming 0 research · 0 corpus · 3 tested - tool · 017
olmOCR AI
Fully open toolkit from the Allen Institute for AI that turns PDFs and scans into clean, reading-order text, built to process millions of pages cheaply.
Free · Some code 0 research · 0 corpus · 5 tested - tool · 019
PaddleOCR-VL AI
Small, free open-weight model from Baidu that reads scans into structured text, tables and formulas in 109 languages. Runs on your own computer.
Free · Programming 0 research · 1 corpus · 1 tested - Project · research
Finnish literary translation circles (opens the project)
Laura Ivaska. Letter networks of Finnish literary translators. Research.
- Project · tested
The Invisibility of Historical Documents in OCR evaluation (opens the project)
Beyene & Dancy (Penn State). Ran OCR tools on The Weekly Advocate (1837), an early African American newspaper. Evaluation.
- Project · corpus
Belgian Historical Gazetteer (opens the project)
Léa Hermenault. Place names from the 1847–55 cadastre. Corpus building.
- Project · tested
Whisper on the Fortunoff Holocaust testimonies (opens the project)
Mattingly & Bailey-Tomecek (Yale). Whisper on 1,847 testimonies in seven languages, scored against human captions. Evaluation.
- Project · corpus
Dictionnaire de Trévoux (1743) OCR pipeline (opens the project)
LIRIS laboratory, Lyon. OCR of an 18th-c. dictionary volume, then LLM correction; also tested Greek and Hebrew. Corpus building.
- Project · corpus
OCR pipeline for historical print periodicals (opens the project)
Marco Lorenz, Research Library for the History of Education (BBF). Open-source batch pipeline for historical periodicals and archival material, pinned to a fixed model version. Corpus building.
- Project · research
Palabras de fidelidad: Bishop González del Campillo’s letters (opens the project)
Maximiliano Abner Alarcón Martínez (Univ. Veracruzana). Distant reading of a bishop's pastoral letters, 1810–11. Research.
- Project · corpus
Reading manuscripts with HTR (opens the project)
University of Pennsylvania Libraries. 17th-c. Italian mathematics manual and 18th-c. Sanskrit text. Corpus building.
- Project · tested
Reading or Guessing? VLM OCR on Greek and Arabic editions (opens the project)
Karamolegkou, Angleraud, Sagot, Clérice (Inria). Compared VLM OCR with Kraken and Tesseract on Greek critical editions and Arabic printed books. Evaluation.
- Project · tested
Reading the Archive by Machine: An OCR Benchmark for Historians (opens the project)
Clifford, Polay, Jack (USask); Humphries, Leddy (Laurier). Benchmarked OCR models on early-modern print, 19th-c. newspapers (incl. Saskatchewan Herald) and Lower Canada manuscripts. Evaluation.
- Project · tested
socOCRbench (opens the project)
Noah Dasanaike (Harvard, Government). Private benchmark of social-science sources incl. pre-1700 and 1700–1950 material. Evaluation.
- Project · corpus
Digitizing 20 million historical census documents (opens the project)
Noah Dasanaike (Harvard). Census images from 9 countries; rejected PaddleOCR-VL-1.5 after fine-tuning problems. Corpus building.
- Project · tested
When Low CER is Not Enough (opens the project)
Gardella, Mariño, Belzarena, Ramírez, Randall, Morel. Tested 18 systems on 175 microfilm pages of 1973–85 military reports (Berrutti Archive). Evaluation.
- Project · tested
Whisper and Portuguese oral history (opens the project)
da Silva, Zagatti, Lopes, Duarte, Bonacin, Alves. Built a local desktop app for oral historians; compared six Whisper models. Evaluation.
- Project · tested
Discovering Object Stories (Recogito Studio) (opens the project)
Middle et al.. Usability-tested Recogito Studio on museum records. Evaluation.
- Project · tested
OCR for Pre-Digital Historical Documents using LLMs (opens the project)
Miller & Banerjee (Virginia Tech). Compared olmOCR with Mistral-Small on degraded early-20th-c. scans. Evaluation.
- Project · tested
ChroniclingAmericaQA (opens the project)
Piryani, Mozafari & Jatowt. QA benchmark of 487,000 question-answer pairs. Evaluation.
- Project · corpus
Indian Affairs letterbooks transcribed with HTR (opens the project)
CRKN & LAC with Gohier, Clifford & Hoy. Transcribed ~600,000 handwritten Indian Affairs letterbook pages. Corpus building.
- Project · corpus
EMEW: a gazetteer of early modern England and Wales (opens the project)
Stephen Gadd et al.. Gazetteer from Adams's Index Villaris (1680). Corpus building.
- Project · research
The gender agency gap in fiction writing, 1850–2010 (opens the project)
Oscar Stuhler. Measured agency of male and female characters in 87,531 works. Research.
- Project · tested
A 13th-century French cartulary in eScriptorium (opens the project)
Michela Galli (Sapienza). Tested layout analysis and HTR models on a 13th-c. French cartulary. Evaluation.
- Project · research
America’s Public Bible (opens the project)
Lincoln Mullen. Found Bible quotations in ~11M pages to argue the Bible's public role. Research.
- Project · corpus
American Stories: structured text of US newspapers (opens the project)
Dell et al.. Re-processed ~20M scans into clean article-level text. Corpus building.
- Project · corpus
Historical Atlas of the Low Countries, 1350–1800 (opens the project)
Rombert Stapel. Locality-level boundary dataset built in QGIS. Corpus building.
- Project · corpus
Captioning archival audiovisual collections (opens the project)
Emory University Libraries. Local Whisper for captioning digitised archival audiovisual collections. Corpus building.
- Project · research
The repression of banditry in Andalusia (opens the project)
Julián Spinelli (Universidad de Buenos Aires). Analysed Zugasti y Sáenz's writings on Andalusian banditry (1876–79). Research.
- Project · corpus
Dutch Global History datasets (opens the project)
Rombert Stapel. Linked datasets tracing Dutch colonial geography. Corpus building.
- Project · tested
NER for the multilingual Persian Gulf (opens the project)
Kapan, Kirmizialtin, Kukreja, Wrisley. Custom NER for British Residency records and Lorimer's Gazetteer. Evaluation.
- Project · corpus
HGIS de las Indias in WHG (opens the project)
HGIS de las Indias. 15,000+ settlements of 18th-c. Spanish America as linked data. Corpus building.
- Project · corpus
BiblIA: medieval Hebrew manuscripts (opens the project)
Stökl Ben Ezra, Brown-DeVost, Jablonski, Kiessling, Lolli, Lapin. ALTO transcriptions of 202 medieval Hebrew manuscript images (BnF, Vatican). Corpus building.
- Project · tested
Cleaning Dirty Books: post-OCR processing (opens the project)
Kim, Pethe, Inoue & Skiena. Deduplication and OCR correction tested on 96,635 texts. Evaluation.
- Project · tested
Digitization Decisions: comparing OCR software (opens the project)
Olson & Berry (Western University), Code4Lib Journal. Tested on a 1928–29 report, a 1975 newspaper and the 19th-c. Voice of the Bondsman. Evaluation.
- Project · research
Feminist Bestsellers: A Digital History of 1970s Feminism (opens the project)
Moravec & Chang. Compared feminist bestsellers against ~40,000 books from 1970. Research.
- Project · research
Burchard of Mount Sion’s Descriptio terrae sanctae (opens the project)
Marianne O'Doherty. Studied the geography of a 13th-c. Holy Land description. Research.
- Project · tested
OCR quality and newspaper clippings (opens the project)
Kettunen et al. (National Library of Finland). Re-OCRed Uusi Suometar (1869–1918) and tested effect on users. Evaluation.
- Project · research
Textile making in the Lienü zhuan (opens the project)
Lu Wang (Western University). Tracked textile terms in biographies of exemplary women, 445–1928. Research.
- Project · research
Sociability in scientific expeditions (opens the project)
Anderson Pereira Antunes. Networks of local collaborators in two 19th-c. Amazon expedition accounts. Research.
- Project · tested
Transcribing the Ottoman Turkish print archive (opens the project)
Kirmizialtin & Wrisley (NYU Abu Dhabi). Custom models for two early-20th-c. Ottoman Turkish periodicals. Evaluation.
- Project · corpus
Enslaved.org: Peoples of the Historical Slave Trade (opens the project)
Enslaved.org. Linked open data platform incorporating Voyages. Corpus building.
- Project · corpus
Geoparsing the Gazetteers of Scotland (opens the project)
Filgueira, Grover, Terras, Alex. Geoparsed 19th-c. Scottish geographical dictionaries. Corpus building.
- Project · corpus
VOC, WIC and notarial deeds: 2 million pages (opens the project)
Liesbeth Keijser, Nationaal Archief (NL). Semi-automatic transcription of ~2 million pages of 17th–19th-c. Dutch company and notarial records. Corpus building.
- Project · tested
Newspaper OCR quality: What Have We Learned? (opens the project)
Lotte Wilms, KB Lab (National Library of the Netherlands). Re-OCR of ~2,000 historical newspaper pages; concluded re-OCR not yet cost-effective. Evaluation.
- Project · corpus
NovelTM datasets for English-language fiction (opens the project)
Underwood, Kimutis & Witte. Identified 210,305 fiction volumes in seven subsets. Corpus building.
- Project · research
Domestic workers of colour in late 19th-century Australia (opens the project)
Caroline Bressey. Recovered domestic workers of colour from 1880–1901 advertisements. Research.
- Project · research
The Many Themes of Humanism in the German press (opens the project)
Hakkarainen & Iftikhar. OCR'd German-language newspapers 1829–50. Research.
- Project · tested
Benchmarking cloud OCR on 180,000 Old Bailey pages (opens the project)
Ughetta & Kernighan. Used transcriptions as ground truth for cloud OCR on 180,000 pages. Evaluation.
- Project · corpus
Deutscher Reichsanzeiger, 1819–1945 (opens the project)
University Library Mannheim. First largely complete OCR of the Deutscher Reichsanzeiger (1819–1945). Corpus building.
- Project · tested
HTR and Sloane’s Miscellanies catalogue (opens the project)
Humbel & Nyhan (UCL). Trained HTR models on Sloane's early-modern catalogue and assessed error rates. Evaluation.
- Project · corpus
Tikkoun Sofrim: Hebrew manuscripts with volunteers (opens the project)
Kuflik, Lavee, Stökl Ben Ezra et al.. Kraken models on four Midrash manuscripts, corrected by volunteers. Corpus building.
- Project · corpus
To be continued: fiction in Australian newspapers (opens the project)
Katherine Bode. Dataset of 9,000+ fiction items found in Trove. Corpus building.
- Project · research
A World of Fiction: Digital Collections and the Future of Literary History (opens the project)
Katherine Bode. Analysed ~10,000 works of fiction in 19th-c. Australian newspapers. Research.
- Project · tested
Ensemble NER for historical place names (opens the project)
Won, Murrieta-Flores, Martins. Place names in the Hartlib Papers and Mary Hamilton Papers. Evaluation.
- Project · tested
Enhanced OCR for mass digitisation (opens the project)
Lars Björk, National Library of Sweden. Combined ABBYY and Tesseract on Swedish newspapers 1818–2018. Evaluation.
- Project · corpus
Relaciones Geográficas maps (opens the project)
LLILAS Benson, UT Austin. StoryMaps of 16th-century Relaciones Geográficas maps. Corpus building.
- Project · research
Telling a tale of World War I with StoryMap (opens the project)
Rebecca Petersen May, Wake Forest. Mapped a WWI soldier's movements from a donated journal. Research.
- Project · corpus
Digital Periegesis (Pausanias) (opens the project)
Foka, Barker et al.. Annotated places and people in Pausanias. Corpus building.
- Project · corpus
Ontario newspapers on a shoestring (opens the project)
Art Rhyno (University of Windsor / OurDigitalWorld). Moved newspaper OCR from ABBYY to Tesseract for microfilmed Ontario papers. Corpus building.
- Project · tested
Arabographic OCR (OpenITI) (opens the project)
Romanov, Miller, Savant, Kiessling. Trained models on printed classical Arabic works. Evaluation.
- Project · research
Gender, discrimination and housing in Montréal, 1901–03 (opens the project)
Robert Sweeny. Linked 1901 census to 1903 property rolls to test landlord behaviour. Research.
- Project · corpus
Old Bailey Corpus 2.0 (opens the project)
Huber, Nissel & Puga. 24M-word linguistic corpus of spoken words in the Proceedings. Corpus building.
- Project · research
Text Mining the Old Bailey for Evidence of Court Behavior (opens the project)
Hitchcock & Turkel. Text-mined all trial reports; argues guilty pleas took hold earlier than thought. Research.
- Project · research
Atlantic History and the Slave Trade to Spanish America (opens the project)
Borucki, Eltis & Wheat. Reassessed size and routes of the trade to Spanish America. Research.
- Project · research
Uncovering reprinted texts in antebellum newspapers (Viral Texts) (opens the project)
Smith, Cordell & Mullen. Text reuse across antebellum newspapers. Research.
- Project · corpus
Finding Names in Trove (opens the project)
Kim & Cassidy. NER over 152M articles; 27M person mentions as linked data. Corpus building.
- Project · corpus
Pelagios 3: early geospatial documents (opens the project)
Simon, Isaksen, Barker & de Soto. 120,000+ place names from pre-1492 maps and texts. Corpus building.
- Project · corpus
A historical commodities lexicon (Trading Consequences) (opens the project)
Klein, Alex & Clifford. ECO as one of five corpora in the Trading Consequences pipeline. Corpus building.
- Project · tested
Rating OCR quality in Early Canadiana Online (opens the project)
Alex & Burns. Scored OCR quality across ~55,000 ECO documents. Evaluation.
- Project · research
Ship crowding and slave mortality (opens the project)
Nicolas Duquette. Linked crowding to mortality after correcting for missing data. Research.
- Project · research
The civilizing process in London’s Old Bailey (opens the project)
Klingenstein, Hitchcock & DeDeo. Compared language of violent and non-violent trials over time. Research.
- Project · research
Coal and place text-mined from Early Canadiana Online (opens the project)
Jim Clifford. Mapped changing geography of coal mentions, long 19th c. Research.
- Project · research
Mining the Dispatch (opens the project)
Robert K. Nelson (University of Richmond). Topic-modelled the Richmond Daily Dispatch, 1861–65. Research.
- Project · research
Visualising the Republic of Letters catalogue (opens the project)
Scott Weingart with Bodleian Digital Library. Visualised 11,284 correspondents in 17th-c. letters. Research.
- Project · research
Topic Modeling Martha Ballard’s Diary (opens the project)
Cameron Blevins. 30 topics from ~10,000 diary entries, charted over time. Research.
- Project · corpus
Montréal, l’avenir du passé (MAP) (opens the project)
Olson & Sweeny. Historical GIS of Montréal released as QGIS applications. Corpus building.
- Project · research
March on Milwaukee digital curation (opens the project)
Jasmine Alinder, UW–Milwaukee. Student exhibitions built around civil rights documents. Research.