View Adaptation in Data Warehousing Systems

An important aspect of a data warehouse system should be its ability to evolve according the user and organization needs. The reason being that the schema of a data warehouse can evolve for integrating new requirements resulting from analysis or data mining processing, schema changes on information sources etc. More precisely, schema changes arising on information sources should be propagated to the related views in order to maintain consistency between the sources and the derived views. In this paper, we present a view model extending the relational view model. And we show how this extended view definition language allows to support view schema changes enforcing consistency and without causing disruptions or modifications to existing views and application programs.