An efficient processing of queries with joins and aggregate functions in data warehousing environment

It is very important to process efficiently expensive queries including joins and/or aggregate functions in a data warehousing environment since there resides an enormous volume of data and the processing of these queries takes a lot of time. In this paper, we propose a new method for processing the queries including both joins and aggregate functions. The proposed method performs grouping of dimension tables with group-by conditions at first and then processes joins by using bitmap join indices. This allows us to process aggregate functions by accessing fact tables only, thus it can reduce the serious performance degradation of existing methods. In order to show the superiority of the proposed method, we develop a cost model for both the proposed and the existing ones, and perform extensive simulations based on the TPC-H benchmark.