Title: Bridging Decision Applications and Multidimensional Databases
Abstract:
Data warehouses were envisioned to provide an architectural model for
the flow of data from operational systems to decision support
environments in order to facilitate reporting and analysis. The
multidimensional models created for decision support systems represent
the conceptual view of the data warehouse the designer has. Typically,
there is an impedance mismatch between this conceptual view and the
physical representation of the multidimensional data. In addition, this
gap between conceptual and physical schemas might cause some
heterogeneity in data instances. We present a framework in which the
relation of
the conceptual model and multidimensional data are specified by a set
of attribute-to-attribute mappings, which are compiled into a set of
views that associate each construct in the conceptual model to a query
on the physical model. Our approach refactors the dimensions in the
conceptual schema and partitions its data instances in order to extract
a set of summarizable dimensions.