Supporting hot spots with materialized views

Since data warehousing has become a major field of research there has been a lot of interest in the selection of materialized views for query optimization. The problem is to find the set of materialized views which yields the highest cost savings for a given set of queries under a certain space constraint. The analytical perspective results in queries which on the one hand require aggregations but oil the other hand arc quite restrictive with regard to the fact data. Usually there are "hot spots", i.e. regions which are requested very frequently, like the current period or the most important product group. However, most algorithms in literature do not consider restrictions of queries and therefore generate only views containing all summary data at a certain aggregation level although the space it occupies could better be used for other, more beneficial views. This article presents an algorithm for the selection of restricted views. The cost savings using this algorithm have been experimentally evaluated to be up to 80% by supplying only 5% additional space.