Data Warehouse Architecture - Evolving Application Domains of Data Warehousing and Mining - page 10

of data warehouse performance and data security
and privacy should be laid behind these data
warehouse services.
Figure 5 depicts example services based on
the prototype architecture in Figure 4. Here we
group the operations and components based on
the major layers of the prototype architecture and
form these different groups into service contracts.
These data-warehouse-oriented service contracts
are listed in the “enterprise service bus” which
contains a complete category of all available
services of an enterprise.
Compared to the prototype architecture de-
picted in Figure 4, SOA-enabled data warehouse
architecture provides benefits such as reusability,
integration, and fast “time-to-market.” Specifi-
cally, a developed data standardization program,
once provided as a service, can be re-used as a part
of the solution to the implementation of other data
quality requirements. The same service can also
be called on by an operational system as part of
its own data validation steps. Integration of data
between the data warehouse and other enterprise

systems becomes easier because different systems
can interact via the same service-oriented platform.
The data warehouse architecture is able to produce
data-as-a-service to end-users and other enterprise
systems. When a business requirement comes, it
is very easy to pick up related data or operations
in the data warehouse services and form them
into the solution. This “mesh-up” concept makes
the implementation of business requirement fast
and agile.
To guarantee these benefits, a service-oriented
data warehouse architecture must contain a uni-
fied meta model and an integrated enterprise
data model. The meta model ensures the smooth
integration of metadata between the data ware-
house and the rest of the enterprise IT systems. An
integrated enterprise data model helps the whole
end-user community of the data warehouse to
have a unified and agreed view of data. Interoper-
ability of the data warehouse metadata and data
guarantees the orchestration of different pieces
in the data warehouse architecture.
A few issues have to be addressed when SOA

Figure 5. Example data warehouse services