Paper proposes MMM data model to improve knowledge interoperability across disciplines and systems

A new paper on arXiv proposes the MMM data model, a normative specification intended to improve knowledge interoperability across disciplines, applications, and deployments. The work argues that traditional document-centric systems constrain how knowledge can be structured, updated, shared, and reused, even as AI reshapes document production. Formal approaches to knowledge representation often prioritize rigid structure over usability, limiting widespread adoption.

The MMM model combines a small set of normative constraints with the expressive freedom of free-text labels. This design aims to enable interoperability without requiring semantic convergence, allowing diverse systems and communities to participate without imposing a single ontological framework. The paper positions MMM within a comparative analysis of the design space of information systems and frames it as a portable alternative to traditional documents for human expression and exchange of knowledge.

To support its claims, the paper reports a reference implementation and pilot deployment data demonstrating implementability and early usability. The author, Mathilde Noual, submitted the paper on June 22, 2026, under the arXiv cs.AI category.