Abstract: OLAP querying is a complex process and the volume of data involved is so large that query response time has become an important warehousing issue. We can select data warehouse views to parttion horizontally in an efficient manner so that OLAP queries will scan fewer fragments and in turn scan fewer rows in the original view. Based on the minterm predicates of OLAP queries, the paper proposes a method for selecting and efficient horizontal partitioning of warehouse cube views in order to improve the efficiency of the OLAP queries.

Key words: data warehouse, materialized views, OLAP queries, metadata