The amounts of data available to decision makers are increasingly important,
given the network availability, low cost storage and diversity of applications.
To maximize the potential of these data within the National Social Security
Fund (NSSF) in Tunisia, we have built a data warehouse as a multidimensional
database, cleaned, homogenized, historicized and consolidated. We used Oracle
Warehouse Builder to extract, transform and load the source data into the Data
Warehouse, by applying the KDD process. We have implemented the Data Warehouse
as an Oracle OLAP. The knowledge extraction has been performed using the Oracle
Discoverer tool. This allowed users to take maximum advantage of knowledge as a
regular report or as ad hoc queries. We started by implementing the main topic
for this public institution, accounting for the movements of insured persons.
The great success that has followed the completion of this work has encouraged
the NSSF to complete the achievement of other topics of interest within the
NSSF. We suggest in the near future to use Multidimensional Data Mining to
extract hidden knowledge and that are not predictable by the OLAP.