We are delighted to announce Session 20: Digital Archaeological Collections as AI Training Data, which will take place at the Computer Applications and Quantitative Methods in Archaeology (CAA) Conference 2026 in Vienna.
📅 Abstract submission deadline: 26 October 2025
🔗 Submit your abstract
📜 Full list of sessions (see S20)
About the Session
Digital archaeological collections hold immense potential as training data for artificial intelligence (AI). This session will explore the opportunities and challenges that emerge when cultural heritage datasets are mobilised for computational purposes.
We welcome contributions reflecting on methodological, technical, ethical, and social dimensions, as well as case studies that demonstrate the possibilities of AI in archaeology.
Suggested Topics
Submissions may address (but are not limited to):
- Case studies on the use of AI in archaeological collections
- Development and exploration of benchmark datasets for AI model testing
- Good practices and guidelines for preparing data files for AI training
- Structuring and curating archaeological datasets in line with Open Science, FAIR, CARE, and TADA principles
- Challenges and limitations of creating or using comparative digital collections for AI
- Technical and ethical issues of bias, representation, and generalisation in AI models trained on heritage data
- Collaborative initiatives supporting open, federated data infrastructures and AI training resources in archaeology
Why Participate?
This session aims to foster critical discussions on how archaeological datasets can be reused for AI in ways that are transparent, reproducible, ethically grounded, and culturally sensitive.
It also strongly connects with the mission of the MAIA COST Action, which promotes the responsible and sustainable integration of AI into archaeological practice. By contributing, you will help shape how data can be mobilised in AI workflows while respecting heritage values and community rights.
Session Organisers
- Vera Moitinho de Almeida (University of Porto & INESCC)
- Nevio Dubbini (University of Pisa)
- Aurore Mathys (Royal Museum for Central Africa & University of Liège)
- Gabriele Gattiglia (University of Pisa)