AHG3 recommendation re ISO 19103

[PeterParslow]

• at present, this describes the use of UML: discuss other CSLs for - other approaches, or perhaps to define a cross-approach CSL
• introduce OWL alongside UML as an alternative CSL and/or create an "OWL profile of UML"
• provide some criteria for choosing between CSLs (at least, UML vs OWL). View: UML can be a good start (even) for a graph database, but usually you need to add the semantics (and a UML profile for doing that would help)
• Update and reinforce the distinction between different levels of abstraction, taking into account the work being done in OGC’s Architecture DWG

See TC 211/N 5561 for more details.

CIB Resolution 2021-02 should start the revision of ISO 19103:2015; the scope (N 5563) currently mentions implementing the learnings from AHG3.

[PeterParslow]

AHG3 report also asked:

Are conceptual schema languages inherently constrained to (or at least better for) particular approaches? Or Is it possible to have one foundational approach that can be implemented in all these ways? Overall, it seems conceivable that a foundational (top level) ontology could then be expressed in both TC 211/OGC UML and Web Ontology Language.

Wikipedia lists seventeen existing "upper ontologies", at least two of which are being considered by the UK's National Digital Twin team. One is published in an ISO standard, but is specific to industrial automation systems (although may have wider application). There is no clear choice for language in which to express such a top level ontology; options include:

• OMG's Meta Object Facility (MOF™) - the meta-metamodel for UML, also used for metamodels for EXPRESS and RDF/OWL
• W3C's draft Manchester Syntax - actually a compact syntax for OWL
• natural language, e.g. English