Call for Papers: Generative AI, Text Mining, and Semantic Modelling – Using Big Models for Big Problems, FAIRly! (CAA2026, Session 33)

—

by

Alphaeus Lien-Talks
in

We are delighted to announce Session 33: Generative AI, Text Mining, and Semantic Modelling – Using Big Models for Big Problems, FAIRly!, 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 S33)

About the Session

Large language models (LLMs), text mining pipelines, and semantic modelling frameworks are transforming the way archaeologists can work with unstructured and complex datasets. But how do we ensure these methods are trustworthy, interoperable, and FAIR?

This session brings together practical applications and critical reflections on combining Generative AI, text mining, and semantic reasoning (e.g., CIDOC CRM, SKOS thesauri, Wikidata) to advance computational archaeology. We welcome contributions that explore end-to-end workflows where semantic knowledge structures guide, constrain, or validate AI outputs, improving their transparency, reproducibility, and reusability.

Suggested Topics

Submissions may address (but are not limited to):

  • Generative AI for text generation, translation, summarisation, or coding in archaeology
  • Text mining and information extraction from grey literature, reports, and multimedia
  • Ontology design, alignment, and knowledge graph construction for archaeology
  • Graph-augmented AI pipelines (e.g., RAG approaches for heritage data)
  • Constraint checking and reasoning with SHACL, OWL, or rule-based methods
  • FAIRification workflows for archaeological AI, including metadata and provenance standards
  • Evaluation metrics, reproducibility protocols, and ethical considerations

Why Participate?

This session highlights the growing need to bridge AI with semantic modelling and FAIR principles in archaeology. It is especially relevant to those interested in developing transparent, schema-aware, and reusable AI pipelines that serve not only research but also long-term heritage data stewardship.

The session builds on the work of the MAIA COST Action, which promotes critical, inclusive, and responsible approaches to AI in heritage.

Session Organisers

  • Alphaeus Lien-Talks (Historic Royal Palaces)
  • Florian Thiery (Leibniz-Zentrum für Archäologie, Mainz & Research Squirrel Engineers Network)