The Pelagios Network

Also, if you are a scholar of the Ancient Near East, consider sending us your stuff! It will be published in our edited collection for Nature:

www.nature.com/collections/...

GitHub - pelagios/llm-lod-enriching-heritage: A repo containing a PythonBook containing recipes created by the Lorentz Center workshop 'Enriching Digital Heritage with LLMs and Linked Open Data' (Leid... A repo containing a PythonBook containing recipes created by the Lorentz Center workshop 'Enriching Digital Heritage with LLMs and Linked Open Data' (Leiden, September 8-12, 2025) - pelagio...

Work is coming together. The GitHub repo is being populated. Follow along here github.com/pelagios/llm...

Blackboard with notes about what the cookbook is for, for whom, and how

Sketching out the users and usability of the LLM infused LOD cookbook — the who, what, how? Courtesy of @miaout.bsky.social & co

The final group working on applications — essentially, what can a cultural heritage professional do with the automated list of disambiguated named entities? But what's going on behind the curtain?

A room showing people working at laptops

A room showing people working at laptops

Meanwhile, work continues furiously on articulating the following two stages of the LLM-LOD pipeline: the reconciliation process itself and then the evaluation of it

Rob Sanderson explaining to Leif Isaksen how our LLM pipeline for preparing reconciliation of named entities is going to work

Rob Sanderson explaining to a thoughtful-looking Leif Isaksen how our LLM pipeline for preparing the automated reconciliation of named entities to authority files (like Wikidata, Pleiades, Nomisma, etc.) is going to work.

Person working at a computer with a screen behind showing computer code

Some real-time coding happening, courtesy of Rob Sanderson

Sketch in chalk of the process for generating candidates for matching NER output to authority files

Day 3 of Enriching Digital Heritage with LLMs + LOD www.lorentzcenter.nl/enriching-di.... From NER to disambiguation: the data group using the latest tech to sketch out how MCPs can generate candidates for linking NER outputs to authority files

In Leiden right now we've just begun a weeklong workshop on exploring the transformative potential of combining Large Language Models (LLMs) & Linked Open Data (LOD) to enrich cultural heritage metadata in ways that foster FAIR usage www.lorentzcenter.nl/enriching-di... Watch this space for outcomes

People Works towards consensus and standards for common representation of historical people in structured Linked Open Data

The "LOD People" activity of the @pelagios.bsky.social Network (pelagios.org/activities/p...) has been collecting samples of data from digital datasets that record historical persons (whether in LOD, TEI XML, CSV, or any other open format.

(Do you have any such data we could look at?)

See bsky.app/profile/abou...

User interface concept for an "essay" feature in #liiive. Images shows the user interface with a new overlay panel containing text about the IIIF object (the 1507 Waldseemüller map). The text includes an embedded annotation snippet image and an annotation link. The viewer (right side) shows the same annotated region as the snippet.

Sketching out an idea for a future #liiive feature:

• Write about a IIIF object
• Embed annotation snippets (links or images).
• Everything stays connected—click a snippet to move the viewer. Maybe even a bit of #scrollytelling?

Curious what you think—would you use this? #DigitalHumanities #IIIF

You can find yourself in this #DH2025 state-of-the-union. In our small way, we are prioritizing "care over speed, sufficiency over spectacle, community over competition", by sharing knowledge, giving advice & building resilience into tool development (thanks to Partners like @performant.bsky.social)

And if you aren’t at #DH2025, ping any of us — happy to chat :)

We’ve been working on all sorts of open source tools, and I am *personally* quite excited about the mapping tools and the annotation tools because, well, #PlaceNerd and #LitNerd. I map and mark things up for fun 😅

Also this from @museologi.st, from a few months back — upping again for those at #dh2025. Dan is a one-man punk cultural heritage operation.

Retweeting with Dan's bluesky handle, @museologi.st, so that he can get the kudos he deserves — and in case this work is of interest to folks at #dh2025

ATRIUM - Recordings wanted for testing the Transcription Portal on picture of archaeologists on site

ATRIUM is calling for Archaeologists to send them audio/video recordings wanted for testing the Speech Transcription Portal in English, Dutch, German and Italian. Help develop this new service! More information: www.ariadne-research-infrastructure.eu/2025/07/18/a...

#TEIFriday
A reminder that if you have #EpiDoc related announcements (projects, jobs, publications, training) that you'd like to announce via Current Epigraphy, contact the EpiDoc rep, and they can circulate it for you. currentepigraphy.org/about/

Preview of the chatbot interface of the "Machina Emblematica" project. A scanned page from the "Symbola Emblematica" is shown on the left, the conversational interface (text, image thumbnails and text input field) is shown on the right.

Chatbot project starting to take shape... #DigitalHumanities #DH2025 #IIIF #LLM

Image annotation with LLMs is a big trend at #DH2025. Work being done across several projects, including our humble poster tomorrow, featuring work on emblem classification and RAG with Michela Vignoli at AIT and @aboutgeo.bsky.social.

Network power: the humanities and data science in collaboration Free event. Registration required.

#DH2025, we're delighted to share a Turing's Humanities and Data Science event in Oxford and online on 25th Sept with a panel asking: 'How far can data science and the humanities help to answer each other’s questions?'

Express your interest here: digitalscholarship.web.ox.ac.uk/event/networ...

The Pelagios Network’s People Activity are holding a Sample Data Gathering Sprint
29 July, 12 - 1pm, Online via Zoom (register for link)
Organisers: Gabriel Bodard (London) and Jun Ogawa (Tokyo)

https://ics.sas.ac.uk/events/linked-open-data-people-data-gathering-sprint

Pelagios is proud to count Performant & Rainer as Partners. Together, we are trying to build methods/tools for annotating & structuring data "in the round", where different Partners contribute knowledge, resource and/or funding, towards distributive resilience. Join our community of practice #dh2025

Sorry to have only found your post now (still getting used to bluesky!). For an e.g. of Recogito in the wild: www.literarygeographies.net/index.php/LitGeogs/article/view/316 NB: There's now the new&shiny recogitostudio.org - fully supported & modular, to enable plugins. Do get in touch if interested

Semantic Annotation without the pointy brackets Work on texts and images. Identify and mark named entities. Use your data in other tools or connect to other data on the Web. Without the need to learn code.

#DH friends: I just came across the Recogito project by @pelagios.bsky.social & have been sharing it with students who are planning #mapping projects building on NER place names.

Nice annotation tool; built-in map; shareable/exportable.

Have any of you been using Recogito?

recogito.pelagios.org

Possibly of interest for folks at #DH2025. Dan is using Rainer's magic liiive.now. Give it a whirl.

Sorry to have only found your post now (still getting used to bluesky!). For an e.g. of Recogito in the wild: www.literarygeographies.net/index.php/LitGeogs/article/view/316 NB: There's now the new&shiny recogitostudio.org - fully supported & modular, to enable plugins. Do get in touch if interested

Performant Studio We partner with scholars in the humanities to build software for research and education.

We’re pleased to announce Performant Studio, an open source, standards-based software toolkit for #digitalhumanities projects. Ask @nicklaiacona.bsky.social about it if you see him at #dh2025! www.performantsoftware.com/studio/

Setting up a IIIF system Instructions on how to set up IIIF services on AWS using Cantaloupe, Directus and S3

The amazing Dan Pett has been at it again. Using @aboutgeo.bsky.social liiive.now Dan explains how to DIY your own IIIF: add image to cms, create manifest, put manifest into Rainer's magic, download annotations, paste text into cms, manifest recreates. Try it yourself! museologi.st/blog/iiif-di...

Historic England Talk - Links This page covers all the links I talked about in my HE conference paper online

The awesome Dan Pett's talk for Historic England on digital innovation in heritage is here museologi.st/talks/he-lin... featuring work for The British Museum, The Portable Antiquities Scheme, The Fitzwilliam Museum, etc. Nearly all the resources he discusses are open source, reproducible & low cost