Lazy Learner Model
The model eventually needs to be "lazy" in considering views in educational content. I am assuming we don't necessarily want to go "backwards" (if someone knows about strings and concatenation, that probably doesn't say much about them knowing about monoids) but going "forwards" should be implemented.
I am also wondering if "lazy" is the correct term here or if this couldn't be just (strictly) propagated, whenever the learner actually encounters the concept of the view itself (i.e. some LO that says "strings with concatenation are a monoid").
In any case, I need to gather more info on how these updates would look like. First resource for that is the relevant paper by Michael and Marcel (see here) and Marcel's proposal (see here).