A data warehouse view selection scheme to accommodate dimension hierarchies.

A data warehouse is a central data repository that supports efficient execution of complex business decision queries. Data warehouse views are aggregations or summary tables holding millions of records integrated from a variety of source data systems. An n-dimensional data cube is a multidimensional data model used to generate n different perspectives of the measure aggregates of interest, and has 2n subviews. Query response time can be significantly improved by pre-computing and storing needed warehouse views. Owing to disk space constraints and increasing maintenance cost of materialized views, pre-computation and storing of all of the required views may not be feasible. Thus, many algorithms have been proposed for selecting only a subset of views of the data cube most beneficial to materialize for better query response time. When taking the warehouse dimension hierarchies into consideration, the view selection problem in a data warehouse gets more complex. The objective of this thesis is to review and contribute a solution to the view-selection problem to accommodate warehouse dimension hierarchies. The proposed selection scheme recommends a set of warehouse views including the dimension subviews with maximum benefits, to materialize in order to improve the query response time in a data warehouse. Source: Masters Abstracts International, Volume: 39-02, page: 0529. Adviser: C. Ezeife. Thesis (M.Sc.)--University of Windsor (Canada), 1998