Aarhus University Seal / Aarhus Universitets segl

Lars Bækgaard

Conceptual Data Warehouse Structures

Publikation: Working paper/Preprint Working paperForskning

  • Erhvervsøkonomisk Institut
  • Information Systems Research Group
We analyze some of the problems related to conceptual modeling in data warehousing environments. The long-term data warehousing challenge is flexibility rather than efficiency. The reason is that most warehouses will contain growing collections of historical data that are supposed to satisfy changing information needs. We show how the event-entity-relationship model (EVER) can be used for schema design and query formulation in data warehouses. Our work is based on a layered data warehouse architecture in which a global data warehouse is used for flexible long-term organization and storage of all warehouse data whereas local data warehouses are used for efficient query formulation and answering. In order to support flexible modeling of global warehouses we use a flexible version of EVER for global schema design. In order to support efficient query formulation in local data warehouses we define a type hierarchy for star schemes that can be used for local schema design. Based on this we outline a query language that can be used to formulate star queries, i.e., queries on star schemes
Antal sider13
StatusUdgivet - 1998

Se relationer på Aarhus Universitet Citationsformater

ID: 204846