Set networks in database applications:
We hypernormalize data from a database as follows:
Each data store has a url down to the database, table, field, and then the data within the field itself. Each is delimited by a ‘!’.
So !!!url!db!table!field!data : this is the absolute symbol.
Thus a table is an association table, associating properties, represented by each of the symbols of the fields, with an object, represented by the symbol for the primary key of the table. All the properties are associated with all of their context with each of the objects with all their context.
!!!context!primary_key:!!!context!data
All of these associations are built into a hypernormalized database. Likewise each symbol contains as a property its raw context itself.
So, !!!context!primary_key:!!!context
This is the basest form of association.
Thus, in context we can query and imply:
|context|property => |another_property
|context|property => object
Implication rules can be discovered by patterns within existing associations. Or they can be explicitly declared and stored in a context.
With logic networks we ought then to be able to generalize rules at the table and field level that apply to granular symbols instantiated from the data itself.
More will come.
