Developing language tools for Ancient Egyptian: my ATRIUM Research Visit at the University of Sheffield
by Roberto A. Díaz Hernández
I am a Junior Lecturer in Ancient History at the University of Jaén in Andalusia, Spain. Since 2024, I have been working on developing Ancient Language Processing tools for Egyptian. My work focuses on developing the Universal Dependencies Egyptian_Pre-Coptic Treebank (EPC treebank). The aim of this project is to annotate the morphological features and syntactic relations of the most representative texts from Old and Middle Egyptian, Classical Egyptian, Late Egyptian and Demotic. The EPC treebank currently contains 34,234 tokens and 3,089 sentences from the Pyramid Texts in its latest release in Universal Dependencies in Mai 2026 (v2.18). The Pyramid Texts form a collection of spells recorded on the walls of the Pyramids of Old Kingdom kings and queens (ca. 2543–2120 BC).
Thanks to the Transnational Access scheme of the ATRIUM programme, I had the opportunity to undertake a nine-week research visit to the University of Sheffield from 20 April to 30 June 2026. During this time, I attended the School of Computer Science daily, where I worked closely alongside colleagues. In a highly stimulating academic environment, I was able to annotate over 5,000 tokens from the Pyramid Texts in the EPC treebank. I also created the EPC parser, a morphosyntactic model trained with the EPC treebank and I made it available for public use on this website:
https://web.ujaen.es/investiga/nile-in-contact/Analizador_en.html
I was happy to become part of academic life in Sheffield, which meant I could attend lectures and seminars organised by Joe Stacey, such as the online seminar called "Understanding the Human-Scale AI Frontier" by Suchir Salhan from the Department of Computer Science at the University of Cambridge. It was an inspiring talk about mechanistic research on small language models. I also had the opportunity to discuss issues relating to the EPC parser with members of Diana Maynard’s research team. This will help me develop it by adding new digital tools in the EPC parser’s website.
I really had a lovely time during my research visit to Sheffield, enjoying not only the exceptional academic environment of its university, but also for its beautiful gardens and green spaces, which are perfect for relaxation and reconnecting with nature. The most beautiful of these places are Sheffield Botanical Gardens. I used to take a walk there after work to watch the amusing squirrels scampering through the trees and to relish the fragrance of the flowers.
A squirrel in Sheffield Botanical Gardens
A sunny day in Sheffield Botanical Gardens
Finally, I would like to thank Diana Maynard for supporting my research. My thanks also go to Joanna Wright for managing all the logistics, and to Joe Stacey for inviting me to his seminar series. I am very grateful to the ATRIUM programme for making this research visit possible, and to everyone at the University of Sheffield’s School of Computer Science who were very kind enough to support my work during this period.
