Some advice I gave for people looking at data modelling technologies:

I would advise them first of all to work on a data model in a language
independent fashion, a bit like how the Atom project started. [So]
they should just chat about the structure of syllabi, with NO
REFERENCE to implementing that model at all. It would be very
important not to bring in any mention of specific data model
implementation technologies, because that can unduly restrict the data
model. Then they should think about how they want to share them and
what they want to do, and HTML [in this case] might be a good fit.

Cf. http://bit.ly/97acCG