Page 134 -

98 Enterprise Data Governance

MDM system knows the transition rules between business
states and uses them to check their modification. For
example, in a system where a business object “product”
cannot move from state “expired” to state “opened”, the
MDM system is the guarantor of this rule and systematically
prevents this update.

A semantic MDM, with the help of business states,
ensures that updates do not corrupt the lifecycle of business
objects. In other words, whatever the organization puts in
place, the states of the objects are in line with the rules
defined by the business. This separation of concerns between
the business layer and the organizational layer is important.
Organizations can evolve while the heart of the way in which
business is carried out stays the same; it therefore becomes
impossible to corrupt the core business even if errors appear
at the level of the organizational processes.

In this way it is possible to manage organizational
variants, without compromising the functioning of the heart
of the MDM system. These variants can co-exist and kick in
according to the organizational context faced (e.g.
headquarter, subsidiary, country, channels etc.); they all rely
on the same core business execution, in a secure and reliable
way.

5.3.3. Preparation for MDM integration with the rest of
a system

Data models have to take into account the requirements
which go beyond those of reference and master data. They
must:

– position the models in an enterprise data architecture
common to that of transactional data;
– facilitate their integration with the rest of the
Information System and its underlying IT systems.