An ants algorithm for optimizing the materialization of fragmented views in data warehouses: Preliminary results

The materialization of fragmented views in data warehouses has the objective of improving the system response time for a given workload. It represents a combinatorial optimization problem arising in the logical design of data warehouses which has so far received little attention from the optimization community. This paper describes the application of a metaheuristic approach, namely the ANTS approach, to this problem. In particular, we propose an integer programming formulation of the problem, derive an efficient lower bound and embed it in an ANTS algorithm. Preliminary computational results, obtained on the well-known TPC-D benchmark, are presented