Skeemis

Skeemis is a theoretical construct in information science and cognitive science that describes a modular representation of knowledge intended to be instantiated across different domains. It emphasizes reusability, composability, and dynamic adaptation as new information becomes available.

The term skeemis is relatively new and lacks a standardized definition. In scholarly discussions it is presented as an extension of the ideas behind schemas and skeemas, aiming to unify declarative knowledge, procedures, and rules within a single structure.

A skeemis model typically consists of nodes representing concepts or entities, edges encoding relations (such as hierarchy, causation, or dependency), and meta-attributes for provenance, versioning, and annotations. The approach supports decomposing large domains into smaller skeemis and mapping between different skeemis to enable interoperability.

Applications appear in adaptive learning systems, where skeemis organize domain knowledge for personalized instruction, in software engineering documentation for modular architectures, and in AI reasoning frameworks that require modular, reusable knowledge bases.

Advantages include clearer knowledge organization, scalability, and transparency of structure. Challenges involve the lack of formal standardization, potential over-specified models, and difficulties in cross-domain evaluation and alignment with existing ontologies.

See also: schema, ontology, knowledge graph.