Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
[mdt-sbvr.dev] Some notes about the MRV implementation

I am starting to experiment with the "Meaning & Representation Vocabulary" (MRV) model, as checked-in for the mdt-sbvr project. These are some initial comments and questions about it.

* Concept.designation and Concept.definition are marked as "derived", so no code is generated for them. And the corresponding getDesignations and getDefinitions methods have not been implemented. What is the planned implementation if these are supposed to be derived? Look through all the Representations in a Package, find those that reference the Concept and are Designations/Definitions? Or build some kind of Map that relates the Concepts to the Definitions or Designations?

* There doesn't seem to a way to navigate from a Text or _expression_ to the corresponding Representation. You need to be able to answer the question: "what Meaning(s) are associated with a given Text?" (Maybe the answer is related to the next one, about Namespaces).

* How do Packages and nested Packages relate to Namespaces and particularly AttributiveNamespaces?

* Looking at the .sbvr file produced, I see that all the cross-references between concepts use XML relative path references. In my experience, such relative references are brittle; they bring if any tag is added or deleted. Granted that we don't expect anyone to edit such a file except with a tool. But I think we should UUIDs or some such mechanism to cross-reference among elements.

* It looks like the Concept.general/generalizes relationship can be navigated in only one direction. I think tools need to be able to go from subtypes to supertypes, and vice-versa. Also, it is not clear whether Concept.getGenerals() returns the subtypes or supertypes of a Concept since "generalizes" and "specializes" are the terms used in SBVR, but not "general". Note that clause 11.1.2 says that the "more general concept" has the narrower intension and the broader extension.

* I don't see support in the model for "placeholder is at starting position" or for "placeholder uses designation".

I'm sure I'll have more questions as I get further into it.

--------------------------------
Mark H. Linehan
STSM, Model Driven Business Transformation
IBM Research

phone: (914) 945-1038 or IBM tieline 862-1038
internet: mlinehan@xxxxxxxxxx


Back to the top