A Method for Data Warehouse Design - Data Warehouse Systems: Design and Implementation - page 400

Table 10.3 Data transformation between sources and the data ware-
house

access information must be specified (e.g., login, password, and accessibility).
Also, for each element in the source schemas (e.g., the entity and relationship
types), we specify its name, its alias, a description of its semantics in the
application domain, and so on. The elements of the data warehouse schema
are also described by names and aliases and, additionally, include information
about data granularity, policies for the preservation of data changes (i.e.,
whether they are kept or discarded upon updates), loading frequencies, and
the purging period, among other things.

10.4.3 Source-Driven Conceptual Design

In this approach, once the operational schemas have been analyzed, the initial
data warehouse schema is developed. Since not all facts will be of interest
for the purpose of decision support, input from users is required to identify
which facts are important. Users can also refine the existing hierarchies since
some of these are sometimes “hidden” in an entity type or a table. As a
consequence, the initial data warehouse schema is modified until it becomes
the final version accepted by the users.
The conceptual-design phase consists of three steps, shown in Fig. 10.8 .
We discuss next these steps.

Validate
conceptual
schema with
users

Define final
conceptual
schema and
mappings

Define initial
conceptual
schema

Fig. 10.8 Steps for conceptual design in the source-driven approach