As information ecosystems grow more heterogeneous, both humans and artificial agents increasingly face a simple yet unresolved question: when seeking knowledge, whom should we ask, and why? Inspired by how people intuitively "read a room", this paper introduces the concept of knowledge affordance (KA) to systematize how agents identify meaningful opportunities for information seeking in hybrid human-AI environments. Rather than introducing a fully formed framework, we propose KAs as declarative, semantically grounded descriptions of what a knowledge source can offer, for which kinds of questions, and with which contextual properties. Additionally, we suggest that KAs are relational, possibly emerging from the interplay between the agent's task, preferences and situational factors. Our contribution is thus a conceptual proposal that connects different research streams, including affordances, semantic web services, knowledge engineering and querying, and mutual intelligibility. We sketch possible research directions to build KA-aware systems that navigate information spaces with greater transparency, adaptability and shared understanding.
Ontologies and Knowledge Graphs are more popular than ever, given their fundamental role in providing context and grounding to (Generative) AI systems. In this landscape, Knowledge Engineering is increasingly framed as a technical methodology to design and automate AI-powered solutions. A key risk of this view is to overlook the management, organizational, and human dimensions of knowledge. Even when encoded as a digital artefact, knowledge remains intrinsically human: it does not become detached from people, but continues to interact with their practices, goals, and interpretations. This tension becomes particularly evident in contemporary hybrid socio-technical systems, where humans and artificial agents, symbolic and non-symbolic AI, and deterministic and autonomous behaviours coexist and interact. This paper argues that Knowledge Engineering needs to be reframed with a human interaction perspective, explicitly leveraging Knowledge Management insights that are often overlooked in current AI-driven settings. By adopting this perspective, new Knowledge Engineering challenges and research questions emerge to better design, evaluate, and use digital knowledge in AI-enabled hybrid socio-technical systems.
Hybrid human-AI systems are increasingly embedded in decision-making contexts, where evaluation is not a single verification step but an interaction-driven process. Existing evaluation approaches, however, often remain fragmented and outcome-oriented, providing a limited view of how evaluation is actually performed and sustained in hybrid collaboration. In this article, we explore meta-evaluation as a means to examine the evaluation process itself in human-AI hybrid systems. We analyse the literature on hybrid evaluation practices, focusing on how evaluation activities emerge within interaction models and workflows, and use existing principles of trustworthy AI as a framework for transparency, reliability, and accountability in evaluation processes. Based on this analysis, we propose a preliminary set of meta-evaluation guidelines intended to support more systematic, explicit, and verifiable evaluation practices in human-AI hybrid contexts.
With digital transformation, industrial companies today are facing the challenges to change and innovate their business, by leveraging digital technologies and tools to support their processes and their operations. One of their main challenges is the management of the company knowledge, especially when tacit and owned by industry workers. In this paper, we illustrate how knowledge graphs can be the turning point to allow industry workers digitize and exploit the knowledge about the “what”, the “how” and the “why” of their everyday activities.In particular, we focus on the “how” by illustrating the challenges related to procedural knowledge management, i.e., the knowledge about processes and workflows that employees need to follow, and comply with, to correctly execute their tasks, in order to improve efficiency and effectiveness, to reduce risks and human errors and to optimize operations. We also explain the relationship in this context between knowledge graphs and sub-symbolic AI approaches.
The advancement of Intelligent Asset Management Systems (IAMS) in the railway sector can be fostered by integrating data and trustworthy artificial/human intelligence. However, the implementation of solutions for maintenance prescription and optimized intervention plans requires the integration of multiple digital artifacts and the involvement of different stakeholders. This paper discusses an IAMS Support Integration Framework, facilitating the integration of diverse digital artifacts for the implementation of intelligent maintenance scenarios in a multi-stakeholder environment. To support the integration, the framework offers functionalities for enhancing data sharing and guaranteeing process tracking within an IAMS. The paper outlines the framework’s requirements and architecture, demonstrates its implementation in practical scenarios from the DAYDREAMS project and presents the preliminary evaluation performed with relevant stakeholders.
