Materialized view selection using exchange function based particle swarm optimization

Data warehousing is an essential part of any effectual business intelligence endeavor. The queries necessary for business decision making against a large data warehouse are usually analytical, complex and exploratory in nature. The facility to answer these queries economically is a critical performance concern in the data warehouse environment. One of the techniques employed in data warehouse to improve query performance is to identify and store the relevant data as summaries or aggregates, referred to as materialized view. The problem of choosing such data and storing them as views has been shown to be an NP-hard problem. This problem has been solved using exchange function based particle swarm optimization (EFPSO) in this paper. Accordingly EFPSO based view selection algorithm (EFPSOVSA) is proposed. Experimentally, it is observed that EFPSOVSA selects comparatively better quality views than the greedy algorithm for view selection.