Hypothetical Queries on Multidimensional Dataset

Hypothetical queries evaluate the effect of adopting or changing some strategy then aiding in decision making. Hypothetical queries simulate some modifications and assess the effect. But in OLAP environments modify data will result in cube recomputation which cost heavily. Through observing and analyzing, hypothetical queries are described as follows: first, dicing the needing hypothetical modified part of cube and modify it then union the remaining part composing hypothetical scenario; second, based on that scenario we calculate result of hypothetical queries. By extending multidimensional algebra and introducing query rewrite, hypothetical queries can be accomplished without actually modify OLAP data. Based on extended multidimensional algebra, two lemma of update as rewrites are concluded. Query rewrite is optimized based on materialized views. We implement hypothetical queries through extending MDX in open source Mondrian OLAP server and experimentally evaluate query rewrite and the overall system.