Making Collections Findable: AI, Metadata, and the Discovery Challenge
Cultural heritage organisations have spent decades building rich, standardised metadata, but are their collections truly discoverable? As AI transforms how users search and interact with digital collections, heritage professionals face a critical question: how do we bridge traditional metadata practices with AI-powered discovery while maintaining the scholarly rigour and contextual depth our collections demand?
This panel explores the intersection of metadata strategy and artificial intelligence in cultural heritage contexts. Panellists will examine how GLAMP organisations are adapting controlled vocabularies, descriptive standards, and cataloguing workflows to support AI-driven search, recommendation engines, and natural language querying…without compromising the precision and authority that heritage metadata requires.
Topics will include:
- Metadata standards and interoperability (LIDO, SPECTRUM, Dublin Core)
- IIIF (International Image Interoperability Framework) and its role in AI-enhanced discovery
- Controlled vocabularies and linked open data (AAT, LCSH, Wikidata integration)
- AI-assisted cataloguing and auto-tagging: opportunities and quality control challenges
- Natural language processing and semantic search in collection management systems
- Balancing machine-readable and human-readable metadata
- Multilingual metadata and cross-cultural discovery
- Ethical considerations: bias in AI systems and inclusive description practices
- Legacy metadata enrichment and remediation strategies
- Measuring discoverability: analytics and user behaviour insights
Moderator: Lisa Grimm, Content Enrichment Lead, Coherent Digital, Co-author, Practical DAM: A Guide for Students and Practitioners
James Mortlock, Senior Digital Archives Manager, HSBC
Tundun Folami, Applications Specialist, Collection, Wellcome Trust