I'm involved in a form of Business Process Modelling we're calling Programme Modelling (as in Programme of Works) (PM). This form of modelling concentrates on the Actors and Artifacts and their relationship to the Activities.
We track the nature of the participation between the actor and the activity. We've created a new diagram type (for our eventual MDG) called a Particigram (after Organogram - the new term for Organization Charts). The Particigram takes an actor as it's root item and uses the relationships to the various behavioural elements to draw the links that it participates in.
Now, as long time readers will know, I'm a big advocate of rendering implicit relationships.
So, an actor (in a Particigram) will have relationships (both explicit and implicit - derived from the explicit relationships) to behavioural elements at various levels (in the case of PM: Phase, Stage, Activity and Action). For example, if an actor is an explicit Decider in an Action, they are an implicit Decider in the enclosing Activity, Stage and Phase.
Within the Repository, the relationships between the behavioural elements is managed by a combination of the browser (parent element) and diagrams (as has been discussed elsewhere).
Elsewhere, I've championed the rendering of these (structural)relationships on diagrams. This is so that when two elements are placed on the same diagram, the nature of their relationship can be visualised.
So, finally, to the question at hand...
How should I render the relationship between a behavioural element and its enclosing elements? It's clearly a meronymy - but should I use Aggregation or Nesting?
I tend to favour Aggregation since an arbitrary element can be a meronym of more than one holonym.
However, I welcome input as to which should be more accurate (or even if another relationship type might be).
TIA,
Paolo
