​The National Information Exchange Model consists of a series of interrelated reference schemas that define all data components found within the various NIEM namespaces, and external namespaces imported into the NIEM reference schema.

The reference schema is a schema for all components within its business context purvie​w. Each of the reference schemas' namespace schema components can also import other namespaces. For example, the NIEM Justice Namespace imports NIEM Core.

There is no support for individual user customizations in the reference schema(s) as reference schemas are intended to support the broadest reuse. Customization occurs in the Information Exchange Package Documentation (IEPD) development process by re-using existing NIEM components, by sub-setting, or extending NIEM where needed components do not exist in the model itself. Customization for a specific type of data exchange culminates in an Exchange schema, which captures the results of sub-setting and extending a schema that articulates the data tags for the information exchange.

These schemas are not part of the reference schema per se, but they are derived from the reference schema or act in conjunction as follows:

• • A Subset schema extracts from reference schemas just those types and elements needed for a specific information exchange. A message instance that validates against the sub-setted reference schemas must also validate against their complete representation in the NIEM model.

  • An Extension schema defines an IEPD-specific namespace to hold types and elements needed for an exchange but do not exist in the NIEM model. Its purpose is to enable reuse of local components and ease IEPD Exchange schema creation.

  • A Exchange schema, as described above, defines the content model of the information exchange.

Each layer of the schemas adds to the ultimate end-to-end interoperability between systems and decreases the cost and time needed for development and implementation. The subset and (exchange) schemas are mandatory for IEPDs.