A Data Warehouse Optimization Strategy Using Binary Chimp for View Materialization

The Materialized View Selection (MVS) problem is a critical NP-complete challenge in data warehouse design, aimed at optimizing query performance while balancing storage and maintenance costs. This paper proposes BChOMVS, a novel metaheuristic approach that utilizes a Binary Chimp Optimization Algorithm (BChO) to efficiently identify near-optimal view sets. Using the Multiple View Processing Plan (MVPP) for view representation, BChOMVS encodes solutions as binary vectors and evaluates them with a comprehensive cost model. Rigorous experimental evaluation on the TPC-H benchmark up to 50GB demonstrates that BChOMVS significantly outperforms state-of-the-art methods like ACOMVS, PSOMVS, and GTMVS. It achieves superior total cost reduction by consistently selecting the smallest, most impactful view sets, establishing itself as the premier choice for large-scale data warehouse optimization where ultimate cost minimization is paramount.