Knowledge Graph Re-engineering Along the Ontological Continuum

A recent paper on arXiv (2605.22093) presents the ontological continuum as a conceptual framework aimed at re-engineering knowledge graphs (KGs) across various modeling methodologies. The authors contend that KGs, essential for contemporary AI and data integration, face challenges due to fragile integration stemming from diverse strategies, ranging from simple vocabularies to complex axiomatised ontologies. This issue is particularly pronounced in neuro-symbolic AI, where the integration of neural and symbolic elements necessitates adapting KGs to meet new demands. Although generative AI provides remarkable automation for this re-engineering, it lacks a solid conceptual foundation without a clear understanding of the KG landscape. The ontological continuum addresses this need, characterized by two orthogonal distinctions: semantics versus pragmatics and properties versus affordances, enabling a framework to describe, compare, navigate, and transform KGs. This paper serves as an expanded version, offering more detail than the initial preprint.