Exploring the frontiers of data science with EIDF
7 October 2025
The Edinburgh International Data Facility (EIDF) hosted its second drop-in session in September, reconfirming its role as a foundation for innovative developments. The session featured presentations from two excellent speakers and dynamic interactions, making it a vital event for professionals and enthusiasts in the data-driven landscape.
Semantic enrichment and geoparsing
The session began with Rosa Filgueira (EPCC Chancellor's Fellow) who discussed her research on semantic enrichment and geoparsing of historical collections from the National Library of Scotland, focusing specifically on the Gazetteers of Scotland.
Rosa highlighted the complexities associated with data formats such as XML and the ongoing challenges posed by OCR (Optical Character Recognition) errors. In an engaging presentation, Rosa demonstrated how she utilises various machine learning models, including advanced Large Language Models (LLMs), to enhance text segmentation and correct errors in scanned documents. She also introduced "frances," an innovative semantic web platform designed to improve data accessibility for non-computer scientists, showcasing the real-world applications and benefits of EIDF resources.
Innovations in wellbeing
Following Rosa’s insightful presentation, Maurice Jamieson (EPCC), who specialises in design informatics, presented his groundbreaking work on the "Era" wellbeing app for the University of Edinburgh Business School.
This app is a cutting-edge solution aimed at helping employees address health issues and receive personalised recommendations. Powered by an LLM model for interactive communication, the app's development and testing are supported by an NVIDIA A100 VM within the EIDF, illustrating the facility's robust infrastructure. Maurice discussed the app's functionalities and its potential to transform workplace wellness by facilitating better personal health management through technology.
Engaging discussions and outlook
After the presentations, the floor opened for a lively discussion in which participants explored the potential and challenges of using LLMs and other technologies for enriching and accessing historical data collections. Attendees raised questions about EIDF services, documentation, and specific technical adjustments, leading to an exchange of knowledge.
After the discussions, Amy Krause (EPCC) shared future developments for EIDF, highlighting upcoming enhancements such as Windows VMs and expanded storage options, again emphasising EIDF's capabilities.
Upcoming drop-in
Don’t miss our next drop-in session on October 21, 2025, at 10:00 AM (UK).
This session will include another lineup of insightful speakers: Alejandro Bordallo (School of Informatics) and Emma King-Smith (School of Chemistry) will explore the realms of autonomous robotics and synthetic organic chemistry, integrating motion planning and machine learning to tackle complex scientific questions.
As data science continues to evolve, EIDF stands as a symbol of learning, exploration, and innovation. Join us next time as we further delve into these exciting frontiers together! Sign up via our EventBrite page.
Stay connected
We look forward to more insightful exchanges at our upcoming session! If you’d like to stay updated on future events or subscribe to our newsletter, please express your interest. You can also tap into EIDF’s extensive resources or create your own application via the EIDF Portal.
Lastly, to watch September's drop-in session, please visit https://edin.ac/46pdYUM.
