An Ant Colony based algorithm for optimal selection of Materialized view

摘要

The problem of Materialized view selection is one key issue in improving response time of complex queries in data warehouse. In this paper we proposed an Ant Colony based algorithm which is efficient compared with Genetic Algorithm (GA), a popular algorithm for materialized view selection. Given space limitation, our algorithm could minimize total query cost as much as possible. Based on evaporation mechanism and positive feedback of Ant Colony Algorithm, our approach could achieve sub-optimal selection in global search area. Experimental results show that our algorithm is much more efficient than GA algorithm with varying space limitations.