Databases Reference
In-Depth Information
into account the states of the business objects. If the state
values are controlled only by IT specialists, in bespoke
software which are inaccessible to the business itself, then
the business administration of the data is sidelined. To
counter this risk it is necessary to look to the next level of
maturity: the semantic MDM system.
5.3. Semantic MDM
A semantic MDM system takes into account all the
richness of the data, in particular the management of the
business objects' states. The idea of semantic modeling has
been looked at in previous chapters and will be looked at
again when covering the procedures associated with
methodology, in the last part of the topic. This approach to
modeling does not satisfy itself with just describing the data;
it is also interested in the definition of validation rules and
in the lifecycles of the objects through their states. Taking
into account this dynamic aspect of the modeling is
unavoidable to design and master all the data knowledge.
This way of modeling is not easy to understand, either for
business users or for IT practitioners. However, it is
necessary to examine it in more depth in order to understand
its importance in the MDM approach.
Why is taking into account the management of business
objects' states complex but unavoidable? First, the states
that are being described here are those which describe the
behavior of the business objects as opposed to the states that
describe the processes in the organization (data approval
processes, workflows). At first sight, this distinction between
the business level and the organization level is not obvious.
Let us take an example to demonstrate the difference. The
business object “Product” follows a lifecycle which can be
described by a series of business states of the type: “opened”,
“suspended”, “outdated”, “archived”, “to be negotiated”.
These states are linked with each other by transitions. For
