Context and challenge

As AI systems become the primary way people and applications access information, the semantic layer (ontologies, knowledge graphs, and contextual metadata) has moved from nice‑to‑have to critical infrastructure. Yet many organisations still struggle to decide when to invest in graphs, how to manage change, and how to align semantic work with data and AI teams.

Meeting the challenge

Time to have some lunch and visit the exhibition. See you back at 14:15 for the start of the afternoon sessions!

Insights from a Recent Survey

Context

Knowledge Organisation Systems (KOSs), such as term lists, thesauri, taxonomies, and ontologies, play a fundamental role in categorising, managing, and retrieving information. 

The first two years of AI hype centred around capabilities. Which AI reasoned best and hallucinated least? As the standard improved and solutions moved on from pilots and experiments to maturity, we’ve realised hallucination isn’t going away any time soon and the magic wands promised by vendors aren’t, you know, magic.

Current achievements and expanding opportunities

For many the transition from General Knowledge LLMs to RAG architectures has been revelationary with benefits including minimal hallucinations, greater accuracy and control over sources. The vast majority of RAG use cases do not use GraphRAG but there is huge interest that continues to grow in using Graphs alongside LLMs and embedding Semantic Guardrails into AI using Ontologies, Taxonomies and other approaches.  

Your chance to visit the DAM Europe and Semantic Data exhibition and network with fellow attendees.

Mapping a database of classical music for use in the music/media industry involves a thorough knowledge of the subject and its many, unique pitfalls. The anatomy of a standard classical music track title, highlighting how it differs from a standard pop music track title and how it contains multiple levels of complex metadata.

In modern eCommerce, product discoverability remains one of the most critical — and often underestimated — drivers of customer experience. This presentation explores how taxonomy and semantic structures can be transformed from a backend necessity into a strategic enabler of findability and scalability. 

With all the hype over algorithms and the steep demand for GPUs, we often forget that AI’s real rocket fuel is data. And it needs lots of it. How much and what data AI companies utilise has turned into a competitive battlefield, with swaths of data becoming walled off and unavailable to others. 

It is ironic that as AI pretends to give us access to unbounded knowledge, behind the scenes an enclosure of knowledge is taking place that is unprecedented in modern history. 

What are the long-term consequences? And if they worry us, what can we do?