As stated in the section on Enterprise Data Management, data, as executed policy is the materialization of the results of the execution of various manual or automated procedures. Data, to be properly reflective of a well-ordered set of enterprise policies must be highly organized such that it is integrated, inter-operable, and non-redundant. Further, the semantics though which data is understood must be well set out.
To achieve all this, data architectures must exist that not only embrace the Enterprise's complete policy domain, but also embrace the full and unfolding life cycle of data. It is well understood that data specifications proceed through multiple levels of generalization. Each level is constructed into one or models that can be employed as reusable specification resources for lower levels of data architecture model development.
The data architecture levels that are constructed from the top to bottom level are Data Element business fact semantics that are, in turn, employed to define business facts within standardized data structures of commonly employed concepts, which, in turn are employed to define database structures, that are, in turn, employed to define Operational Databases that enable the capture and storage of actual enterprise data.