Databases Reference
In-Depth Information
far removed from the precise description of data and
associated validation rules that are, in fact, required. At best
they express generalities which establish the high level aims
of what the future system might look like, with sometimes
detailed rules, but which cover only part of the requirements.
In most companies, these models contain data duplication,
are ambiguous and incomplete. IT specialists cannot use
them exactly as they are to design programs. Instead, they
have to complete them and distort them, which lead IT
practitioners to detail what should have been done at the
business level. Thus, IT teams try to replace their business
users without having either the vocation or the skills.
The MDM approach cannot allow such a breakdown
between the requirements and the build of the software. The
alignment of the governance functions with the business
requirements will not happen if this breakdown is not
removed.
It is necessary to increase the modeling effort conducted
by the business to support a repository driven by models, i.e.
a Model-Driven MDM. Commitment to this model must be
enshrined in the highest levels of the business management
and, since MDM acts on data shared between different
business units, it is necessary that the organization
promotes the sharing of models.
But business users alone are not able to supply sufficient
effort to achieve such an ambition in relation to modeling. If
each business unit within a company designed its own data
model, then the company would be faced with as many data
models. Only executive management can supply sufficient
impetus to demand the co-ordination required for the work
involved in data modeling at the Information System level.
