Page 51 -

12 Enterprise Data Governance

It would be best to include or make clear these rules in
5
data models . We therefore avoid the risks associated with
an approximate modeling that has to be completed by
specifications found outside the models which are hard to
maintain and understand. This is the reason why rich data
models are needed. It is not just about a static description of
information. Modeling also takes into account a dynamic
description of the data, i.e. the validation rules, the behavior
of associations depending on the use contexts and the
lifecycles of the business objects. In order to obtain this
6
model, we will see that semantic modeling procedures must
be applied.

1.3. Reference/Master data definition

There is no standard definition for reference and master
data, although attempts have already been made, such as:

– “Among all data, some is more critical for the
business activity and the IT system as they are for the most
part shared between a number of applications: we shall call
it reference and master data” [REG 08];
– “Sometimes called reference data, master data consist
of facts that define a business entity – facts that may be used
to model one or more definitions or views of an entity. Entity
definitions based on master data provide business
consistency and data integrity when multiple IT systems
across an organization (or beyond) identify the same entity
differently” [RUS 08];

5. Simple rules are directly included in the data model; the more complex
rules are declared and their implementation is based, for preference, on a
business rules management system.
6. These are described in detail in the last part of this book. They are not
familiar to the management world, nor to business practice. Nonetheless,
we strongly recommend a careful reading of Chapter 8 to acquire the basis
for this new approach of the formalization of knowledge.