More Semantics Please

As I suspected, while I know much of the Python syntax from using the 2.5/2.6/2.7 versions with a major project and several assignments now, there's likely to be a lot of semantic learning and discovery on my part.

For example, OPUS provides Gridcell, Parcel and Zone based models in the 4.3.0 Windows release. My understanding so far is this means in at least one context (based on Chapter 13 of the monster 300-page OPUS User Guide and informal discussions with the NSW Bureau of Transport Statistics):

• Gridcell - arbitrary grid squares/rectangles chosen, no direct/intuitive mapping to most real-world data sets.
• Zone - Typically an aggregation of parcels, often used in Transport Models.
• Parcel - An aggregation of buildings (which are aggregations of households), i.e. much more direct/intuitive mapping to real-world household locations.

Of course, OPUS is not limited to just this. It appears that further arbitrary (but I'm sure well-intentioned) model constructs such as neighbourhoods, regions and "faz" (most likely an acronym for some other kind of land group or classification from the source code in \opus\src\psrc\faz ) also exist. Indeed from Chapter 17.1.3 of the OPUS User Guide,

"For example, an application in Paris used Communes, or
administrative areas roughly equivalent to traffic analysis zones in number (approximately 1300)."

What does that tell me? If nothing else, OPUS is extensible so the questions are more likely to become purely semantic and thus limited by our collective understanding of the world:

what's the best known way to model this concept or idea given the current limits of computing and our ability as human beings to bridge the syntactic and semantic gaps?

rather than something like:

how do I get my city's data into OPUS format?

Of course, there's a long way to go yet. These concepts apart from the "household" or "building" level may have no real-world analogue.

Literally - if you needed to split a city up into parcels, zones or neighbourhoods, how would you do it?

Perhaps more importantly how would you figure out how Australian concepts (not sure how applicable they are worldwide) like suburbs, postcodes, electorates and regions (just to take a few possibilities) map onto these OPUS constructs?

Having worked in government I could see questions from a minister like, how is the land use in my electorate going to change over the next 30 years? What kind of projects should my urban planners be pursing today? What options should we be considering to achieve the best short and long term outcomes for my community given known budget constraints?

And to begin to answer such kinds of questions, I need more semantics. Study ho!