Four iCoSys researchers contributed to a hands-on tutorial on knowledge graphs and agentic AI at the International Semantic Web Conference (ISWC 2026), held on 25–26 October 2026.

Our colleagues Ornella Vaccarelli, Olivier D’Ancona, Anna Scius-Bertrand and Beat Wolf, together with Federica Bove from the HumanTech Institute, presented ARGO: Agentic Retrieval and Graph Orchestration for Document Knowledge Systems, a half-day tutorial exploring how semantic technologies and AI agents can be combined to build more transparent and reliable knowledge systems.

From information retrieval to knowledge-driven AI

How can AI systems go beyond retrieving information to provide answers grounded in verifiable evidence?

The ARGO tutorial addressed this question through a practical introduction to semantic extraction, knowledge graph construction, GraphRAG and agentic orchestration.

Participants explored how to extract structured information from documents and web sources, organize it into knowledge graphs, retrieve relevant evidence and integrate these capabilities into AI agent workflows.

A particular focus was placed on provenance, trust and explainability, showing how explicit links between sources, entities and relationships can support more transparent and auditable AI-generated answers.

A hands-on approach

The tutorial combined theoretical foundations with practical exercises, guiding participants through the main components of a web knowledge system.

Using ARGO’s open-source materials, participants could explore how graph-based retrieval and AI tools work together, as well as the challenges involved in maintaining reliable and traceable information throughout the process.

Congratulations to the team!

Congratulations to Ornella, Olivier, Anna, Beat and Federica for their contribution to ISWC 2026 and for sharing their expertise with the international Semantic Web community.

Their involvement highlights iCoSys’s activities at the intersection of artificial intelligence, knowledge representation and trustworthy AI.

More information: https://argoiswc.github.io/