Page 65 -

26 Enterprise Data Governance

The next maturity level entails putting an MDM system
in place in front of the silos in order to create, in one
location, the true version of reference and master data. This
level attempts to get there without worrying about semantic
modeling. MDM revolves here around a database that has a
similar structure to the physical data models already in
place. Even though this MDM system is not the one we
recommend, it still adds to the quality of the data. It takes
into account time, context and version management. At this
stage, the maturity level is qualified as “extended quality”. It
is important to pay attention to this approach, because the
capacity of MDM to evolve is generally not sufficient:
without rich and stable data modeling, the company runs the
risk of creating MDM systems through silos.

Level 3 – TOTAL DATA QUALITY
• Data repository based on a rich data model independent of
existing databases schemas. It brings all data governance
MDM
functions aligned with the semantic data modeling

Level 2 – EXTENDED DATA QUALITY
Data Steward • Data repository with limited data governance
functions as it does not rely on a semantic data
MDM model. Its data model has a similar structure to
existing databases schemas

Level 1 – BASIC DATA QUALITY

• Sticking cleansing tools
on existing silos and databases
• Lack of data governance

Figure 1.4. The level of maturity of data quality

Finally, the highest level of maturity is based on MDM
taking advantage of a semantic data model. This rich data
model is independent of the physical schemas of the
databases; it expresses all the data meanings, with their