I have put up a new wiki page at
http://www.openehr.org/wiki/display/spec/Towards+a+definition+of+'slot'+semantics
on how to better understand slot semantics. The idea is to a) fix some
existing problems in archetypes and b) get as much as possible of the
template semantics into the ADL/AOM formalism.
reactions welcome.
- thomas beale
Regarding regular expression inclusion check, I think it's not *that*
hard to do. Patterns are checked using purely finite deterministic
automata operations. The computational cost is the cost of generating
the minimum automata (and the automata is built from the regular
expression). If you want to know more I can recheck the algorithm for
you (I implemented this like 3 or 4 years ago and I don't have right
here the process).
I like the proposed way of matching concepts in slots, although I
think that the problem of matching two codes of two distinct
terminologies/ontologies is way much harder.
Regarding the "human understanding" of the ADL slots, IMHO it's not a
big issue: If someone is "brave" enough to open and understand an ADL
file, then he will be able to understand the slots as they are. Tools
should be the ones to soften the edge between ADL and the meaning.
Said that, I'm for the proposed changes (most of them)