Materialized View Selection Using Bumble Bee Mating Optimization

Materialized View Selection Using Bumble Bee Mating Optimization

Biri Arun (School of Computer and Systems Sciences, Jawaharlal Nehru University, New Delhi, India) and T.V. Vijay Kumar (School of Computer and Systems Sciences, Jawaharlal Nehru University, New Delhi, India)
Copyright: © 2017 |Pages: 27
DOI: 10.4018/IJDSST.2017070101
OnDemand PDF Download:
List Price: $37.50
10% Discount:-$3.75


Decision support systems (DSS) constitute one of the most crucial components of almost every corporation's information system. Data warehouse provides the DSS with massive volumes of quality corporate data for analysis. On account of the volume of corporate data, its processing time of on-line analytical queries is huge (in hours and days). Materialized views have been used to substantially improve query performance. Nevertheless, selecting appropriate sets of materialized views is an NP-Complete problem. In this paper, a new discrete bumble bee mating inspired view selection algorithm (BBMVSA) that selects Top-K views from a multidimensional lattice has been proposed. Experimental results show that BBMVSA was able to select fairly good quality Top-K views incurring a lower TVEC. Materialization of the selected views would improve the overall data analysis of DSS and would facilitate the decision making process.
Article Preview

1. Introduction

Decision making is one of the most crucial activities undertaken by individuals, business corporations, governments, etc. In order to assist managers of all levels, decision support systems (DSS) were developed, which, presently, have become one of the most essential components of almost every corporation's information system. It is a computer system, which data managers and analysts negotiate to properly analyze corporate data and acquire the past and present business information and knowledge therein for the purpose of supporting managers in strategic decision making in real time (Sauter, 2010; Turban et al., 2005). The quality of the corporate data, which is processed by the DSS, to provide information to analysts, is of supreme importance alongside analytical models and algorithms. Poor quality of data would cause DSS to yield inaccurate, erroneous and misleading information, which when acted upon, would adversely affect corporation's business. Corporations lose hundreds of billions of dollars annually due to poor quality of data all over the world (Turban et al., 2005).

In order to ensure high quality corporate data, data warehouse was developed. A data warehouse extracts dirty data from various transactional databases of various departments, and transforms and integrates it to produce a single uniform version of the corporate data. This is then loaded into the central database of the data warehouse (Jensen et al., 2010; Inmon, 2003). Data that enters into the data warehouse is never deleted, but is archived throughout the life of the data warehouse. Its data is never updated; instead it stores every update of the data. Unlike transactional databases, data in data warehouse is not stored based on business processes, but is stored based on business subjects. Data warehouse is the framework that provides integrated, historical, non-volatile, subject-oriented and time-variant corporate data, which is highly reliable, to the DSS of the corporation (Jensen et al., 2010). The database of the data warehouse is designed for fast retrieval of huge volumes of data as required during data analysis. Unlike transactional databases, it stores data using the multidimensional data model in the form of de-normalized structures. The volume of its database grows with time and its tables become massive in size. On-line analytical queries posed against such massive base tables of the data warehouse require enormous amount of processing time, in hours and days, to process and extract the desired information; but its required response time is of few seconds (Gupta et al., 1997). Increased response times, degrade on-line data analysis, due to which the extracted information may become unsuitable for making any competitive and strategic business decision. The performance of on-line analytical queries can be improved using materialized views, indexing, join indices etc. (Jensen et al., 2010). This paper focuses on the use of materialized views to improve the performance of on-line analytical queries, which in turn would facilitate better corporate decision making.

Complete Article List

Search this Journal:
Volume 15: 1 Issue (2023): Forthcoming, Available for Pre-Order
Volume 14: 4 Issues (2022): 1 Released, 3 Forthcoming
Volume 13: 4 Issues (2021)
Volume 12: 4 Issues (2020)
Volume 11: 4 Issues (2019)
Volume 10: 4 Issues (2018)
Volume 9: 4 Issues (2017)
Volume 8: 4 Issues (2016)
Volume 7: 4 Issues (2015)
Volume 6: 4 Issues (2014)
Volume 5: 4 Issues (2013)
Volume 4: 4 Issues (2012)
Volume 3: 4 Issues (2011)
Volume 2: 4 Issues (2010)
Volume 1: 4 Issues (2009)
View Complete Journal Contents Listing