Where: Pegaso

URL: https://mkg.infinity-eccch.eu

Mainstream Knowledge Graphs (KGs) often assume a monolithic notion of truth, offering limited support for context-dependent validity, conflicting viewpoints, knowledge evolution, and concept drift. This full-day Barstuhl workshop introduces emph{Multi-dimensional Knowledge Graphs (MKGs)} as an umbrella paradigm for contextual, perspectival, and temporal knowledge representation. It will bring together formal and engineering perspectives to examine key modelling formalisms, scalability challenges, and reference use cases through lightning pitches and collaborative breakouts. The main outcome will be a community-driven framework document that will serve as the foundation for an MKG white paper and a potential future Dagstuhl seminar.

Where: Orione

URL: https://dhlab-nl.github.io/charged-workshop/

Data is the key to understanding and solving challenges such as the loss of biodiversity, predicting recurring weather phenomena, or the fragmentation and loss of context of our cultural heritage. For this, vast amounts of data from many different sources need to be interpreted. Due to advances in computer power and the increasing availability of data spanning longer periods of time, historical data is becoming more important to include. Unfortunately, many AI tools are not well-suited to deal with this type of data and integrating contemporary and historical data streams is a complex problem. In this workshop, we bring together computer scientists, digital humanities researchers, and cultural heritage professionals to create a whitepaper detailing the challenges for AI-assisted integration of heterogeneous and asynchronous data streams from contemporary and historical sources.

• Reason 1

• Reason 2

• Reason 3

Where: Pegaso

URL: https://redq-workshop.github.io

Knowledge Graphs (KGs) are increasingly relied upon as semantic infrastructure for AI systems — grounding generative models, supporting retrieval, and serving as machine-actionable memory for autonomous agents. These roles place new demands on KGs that traditional data quality (DQ) dimensions — accuracy, completeness, consistency, timeliness — only partially capture: what matters is no longer just whether facts are correct, but whether a KG’s semantics are sound enough for an AI system to depend on, including well-formed ontological commitments, correctly disambiguated entities and relations, traceable provenance, and machine-interpretable meaning. A KG can score well on conventional DQ metrics and still fail as AI infrastructure, since classical frameworks were designed for static, human-facing data rather than knowledge that machines interpret, traverse, and reason over. This Dagstuhl-style workshop argues that AI usage surfaces a new DQ landscape specific to KGs as semantic artifacts, distinct from DQ for AI in general, and brings together researchers and practitioners from the Semantic Web, Knowledge Graph, data management, and AI communities to identify the DQ dimensions distinctive to KGs in AI contexts and develop a shared research agenda going forward.

Where: Orione

URL: https://sawa-workshop.github.io

The emergence of LLM-based agentic AI has recreated many of the challenges formally faced by the Semantic Web Services (SWS) community two decades ago: how should agents with differing ontological commitments discover, compose, and invoke capabilities coherently? The original SWS response provided formally grounded capability descriptions and mechanisms for bridging ontological mismatches. Current agentic AI systems largely lack this formal rigour, reintroducing well-understood failure modes, such as a reliance on linguistic descriptions of services/tools/capabilities without enforcing semantic guarantees. Furthermore, the Multi-Agent Systems (MAS) community has developed rich models of agent-environment interaction through affordances, formalising what an agent can do given its current state and the environment it inhabits. This discussion-led workshop will re-examine the role of semantics within the description of agents and services, with a specific focus on the affordance frameworks that cut across the MAS, SWS, and Agentic AI communities. Through discussion, we will critically evaluate how this research, together with the use of the Semantic Web can inform the design of LLM-enabled agentic systems. The result will identify key challenges, open problems, and a shared research agenda for semantically grounded, interoperable agentic AI.

• Reason 1

• Reason 2

• Reason 3

Where: Orione

URL: https://taapaai.github.io

The emergence of LLM-based agentic AI has recreated many of the challenges formally faced by the Semantic Web Services (SWS) community two decades ago: how should agents with differing ontological commitments discover, compose, and invoke capabilities coherently? The original SWS response provided formally grounded capability descriptions and mechanisms for bridging ontological mismatches. Current agentic AI systems largely lack this formal rigour, reintroducing well-understood failure modes, such as a reliance on linguistic descriptions of services/tools/capabilities without enforcing semantic guarantees. Furthermore, the Multi-Agent Systems (MAS) community has developed rich models of agent-environment interaction through affordances, formalising what an agent can do given its current state and the environment it inhabits. This discussion-led workshop will re-examine the role of semantics within the description of agents and services, with a specific focus on the affordance frameworks that cut across the MAS, SWS, and Agentic AI communities. Through discussion, we will critically evaluate how this research, together with the use of the Semantic Web can inform the design of LLM-enabled agentic systems. The result will identify key challenges, open problems, and a shared research agenda for semantically grounded, interoperable agentic AI.