Databases Reference
In-Depth Information
The English phrase applied to the MDM field expresses
this quite well: “one version of the truth”. But this simple
explanation is not always entirely achieved, for two main
reasons:
-
the first is the result of data duplication in multiple
systems. In the event of conflict between different values,
from one database to another, a quality problem appears.
How do we find the correct value?
-
the second reason, less obvious, takes into account the
new forces that are in play in IT systems. These forces
appear along three lines: time management, context
management and version management. Data must be
positioned in this three-dimensional space, as it is otherwise
impossible to govern its value (Figure 1.3).
1.4.1.1
.
Time management
A single piece of data can have a different value over time,
be it in the past or in the future. For example, an
organization can update data for a future date in order to
prepare a regulation change. This time constraint is rarely
taken into account sufficiently in existing systems, even
though it is needed more and more. This requirement, which
certain industries are already aware of, especially insurance
companies under regulation pressure, is becoming more and
more generalized. It is the corollary of the increase in the
expectation of process traceability. Certain pieces of data,
particularly sensitive to the time factor, are accompanied by
periods of validity, enabling the process to acquire the
correct value depending on the execution date.
Taking this into account is not always easy, especially
when it concerns date validation management on links
between data. For instance, an organization describes the
relationship between a
product
and a
factory
for a given
period of time, and then switches
factory
for another period.
