A Multi-Objective Approach for Materialized View Selection

In today's world, business transactional data has become the critical part of all business-related decisions. For this purpose, complex analytical queries have been run on transactional data to get the relevant information, from therein, for decision making. These complex queries consume a lot of time to execute as data is spread across multiple disparate locations. Materializing views in the data warehouse can be used to speed up processing of these complex analytical queries. Materializing all possible views is infeasible due to storage space constraint and view maintenance cost. Hence, a subset of relevant views needs to be selected for materialization that reduces the response time of analytical queries. Optimal selection of subset of views is shown to be an NP-Complete problem. In this article, a non-Pareto based genetic algorithm, is proposed, that selects Top-K views for materialization from a multidimensional lattice. An experiments-based comparison of the proposed algorithm with the most fundamental view selection algorithm, HRUA, shows that the former performs comparatively better than the latter. Thus, materializing views selected by using the proposed algorithm would improve the query response time of analytical queries and thereby facilitate in decision making.

MVPP-Based Materialized View Selection in Data Warehouses Using Simulated Annealing

The process of extracting data from different heterogeneous data sources, transforming them into an integrated, unified and cleaned repository, and storing the result as a single entity leads to the construction of a data warehouse (DW), which facilitates access to data for the users of information systems and decision support systems. Due to their enormous volumes of data, processing of analytical queries of decision support systems need to scan very large amounts of data, which has a negative effect on the systems’ response time. Because of the special importance of online analytical processing (OLAP) in these systems, to enhance the performance and improve the query response time of the system, an appropriate number of views of the DW are selected for materialization and will be utilized for responding to the analytical queries, instead of direct access to the base relations. Memory constraint and views maintenance overhead are two main limitations that make it impossible, in most cases, to materialize all views of the DW. Selecting a proper set of views of DW for materialization, called materialized view selection (MVS) problem, is an important research issue that has been focused in various papers. In this paper, we have proposed a method, called P-SA, to select an appropriate set of views using an improved version of simulated annealing (SA) algorithm that utilizes a proper neighborhood selection strategy. P-SA uses the multiple view processing plan (MVPP) structure for selecting the views. Data and queries of a benchmark DW have been used in experimental results for evaluating the introduced method. The experimental results show better performance of the P-SA compared to other SA-based MVS methods for increasing the number of queries, in terms of the total cost of view maintenance and query processing. Moreover, the total cost of queries in the P-SA is also better than the other important SA-based MVS methods of the literature when the size of the DW is increased.

Random Walk Grey Wolf Optimizer Algorithm for Materialized View Selection (RWGWOMVS)

Optimal selection of materialized views is crucial for enhancing the performance and efficiency of data warehouse to render decisions effectively. Numerous evolutionary optimization algorithms like particle swarm optimization (PSO), genetic algorithm (GA), bee colony optimization (BCO), backtracking search optimization algorithm (BSA), etc. have been used by researchers for the selection of views optimally. Various frameworks like multiple view processing plan (MVPP), lattice, and AND-OR view graphs have been used for representing the problem space of MVS problem. In this chapter, the authors have implemented random walk grey wolf optimizer (RWGWO) algorithm for materialized view selection (i.e., RWGWOMVS) on lattice framework to find an optimal set of views within the space constraint. RWGWOMVS gives superior results in terms of minimum total query processing cost when compared with GA, BSA, and PSO algorithm. The proposed method scales well on increasing the lattice dimensions and on increasing the number of queries triggered by users.