| Previous | Table of Contents | Next |
Data warehousing is a recent development in enterprise-scale information management. The data warehouse technique recognizes
that it is impractical to manage the information of an enterprise as a unified logical database. Instead, this technique extracts
information from logically- and physically-distinct databases, integrates the information, and stores it in a large-scale
“warehouse? database that allows read-only access to possibly non-current data. The extraction and integration processes depend
on a database administrator creating a mapping from the schemas for the individual databases to the schema of the warehouse.
If the meta-information for the various databases is represented using MOF-based technology, then it should be possible to
create sophisticated tools to assist the database administrator in this process.
Meta data is often described as the “heart and soul? of the data warehouse environment. The MOF can be used to automate meta
data management of data warehouses. Current meta data repositories that manage data warehouses often use static meta data
using batch file-based meta data exchange mechanisms. We expect the use of MOF- and standard CORBA-based event and messaging
mechanisms and mobile agent technology (also being standardized by OMG) to drive a new generation of data warehouse management
tools and systems that are more dynamic. These tools will enable customers to react in a timelier manner to changing data
access patterns and newly discovered patterns, which is the focus of data mining and information discovery systems.
The MOF interfaces and the MOF Model can be used to define specific metamodels for database, data warehouse, model transformation,
and warehouse management domains. The integration between these models in a run time data warehouse and the development environment
(which has data models) and UML based object models (which describes the corporate data models and operational databases)
is a typical use of a MOF. The traceability across these environments is enabled by defining an impact analysis metamodel,
which builds on the rich model of relationships supported by the MOF.
The OMG CWM specification represents one possible realization of this scenario.