KG-NeSy 2026: The Third International Workshop on Knowledge Graphs and Neurosymbolic AI

Knowledge Graphs (KG) provide structured and machine-interpretable knowledge that supports integration, reasoning, and explainability. Their combination with machine learning has become increasingly important within Neurosymbolic AI, which integrates neural and symbolic approaches to combine data-driven learning with knowledge representation and reasoning.

The Knowledge Graphs and Neurosymbolic AI Systems (KG-NeSy) workshop focuses on this rapidly growing combination between KGs and Neurosymbolic AI, highlighting how these two research areas can mutually enhance each other. In particular, the workshop addresses the following complementary strands of research:

• KGs within Neurosymbolic (NeSy) AI systems, either as (a) deeply integrated components supporting hybrid reasoning, as (b) auxiliary structures improving explainability, interpretability, robustness, generalization, and transferability of deep learning models, or as (c) structured inputs enabling neurosymbolic prediction and inference.
  This direction is motivated by the observation that, while deep learning excels at processing raw data, it often struggles with planning and deductive reasoning—capabilities provided by symbolic knowledge representations.
• NeSy AI to support KG engineering. Neurosymbolic AI approaches can assist within Knowledge Graphs engineering lifecycle, e.g., within ontology construction, entity and relation extraction, knowledge integration and mapping, completion through logical reasoning or validation, refinement via learning-based methods, and continuous evaluation and maintenance.
  Even when KGs are not explicitly embedded inside Neurosymbolic AI approaches, these approaches can improve methods and tools for knowledge engineering, paving the way for the next generation of Knowledge Graphs development practices.
• KG-based approaches for understanding and systematizing NeSy AI, e.g., through design patterns, taxonomies, conceptual frameworks, or boxology representations. These works are crucial for identifying gaps in current approaches, aligning fragmented research landscapes, and opening up new opportunities for future developments.

Topics

KGs within Neurosymbolic (NeSy) AI systems

• Knowledge Graphs for trustworthy NeSy AI systems.
• Symbolic methods for Explainable AI.
• Hybrid reasoning combining Deep Learning with Semantic Web.
• KG-based approaches for improving fairness/bias of ML algorithms/systems.

NeSy AI to support KG engineering

• NeSy AI methods/systems for ontology learning, construction, or population.
• NeSy AI methods/systems for knowledge extraction, integration, and completion, evolution/ maintenance, and evaluation.

KG-based approaches for understanding and systematizing NeSy AI

• Meta-models capturing symbolic-subsymbolic interplay.
• Formal representations, boxology, design patterns, and taxonomies of NeSy architectures.
• Metrics and datasets to evaluate NeSy AI.

Application of KGs and NeSy AI

• Applications of NeSy AI in domains such as medicine, biology, manufacturing, IoT, security, robotics, education, and others.
• Certification and auditing of AI systems using Knowledge Graphs and NeSy AI.
• The interplay between Knowledge Graphs, NeSy AI systems, and human intelligence.

Authors Guideline

We welcome the following types of contributions:

• Full paper (12-15 pages)
• Short paper (5-8 pages)
• Extended abstract from published papers (2-4 pages)

All submissions must be written in English and adhere to the CEUR-ART style (one column). Please use the following template.

We follow a single-blind process with at least two reviewers per paper. Papers will be evaluated according to their significance, originality, technical content, style, clarity, and relevance to the workshop.

Please submit your contributions electronically in PDF format via the EasyChair system through the submission link.

Accepted contributions will be presented at the workshop and included in the CEUR workshop proceedings. At least one author of each article is expected to register for the workshop and attend to present their contribution. For any enquiries, please send an email to: kgnesy2026 [at] easychair [dot] org.