Fragmentation in Distributed Database Design Based on Ant Colony Optimization Technique

Fragmentation in Distributed Database Design Based on Ant Colony Optimization Technique

Van Nghia Luong, Vijender Kumar Solanki, Nguyen Ha Huy Cuong
Copyright: © 2019 |Pages: 10
DOI: 10.4018/IJIRR.2019040103
(Individual Articles)
No Current Special Offers


Distributed database design solutions depend heavily on the exploitation of input data sources by using clustering techniques in data mining. A new approach of biomimetic computation systems such as ant colony optimization (ACO) for this solution is of interest to informatics experts. Using ACO techniques for this solution has the advantages such as faster algorithms thanks to the randomness of ant colony behavior. The use of random numbers based on heuristic information to pickup (drop) points will facilitate the flexible search on a large data space, so that it provides us with a better answer. In this article, the authors present ACO algorithms application solutions to clustering techniques for the problem of vertical fragmentation of distributed data.
Article Preview

1. Introduction

Ant Colony Optimization method is used to design heuristic for the algorithm that solves combinatorial optimization problem as mentioned above (Dorigo & Di Caro, 1999). The first algorithm was proposed in 1991 by M. Dorigo and then there have been many versions based on the fundamental principles of Ant Colony Optimization discussed in the scientific conferences, information technology projects after the 90s.

Ant Colony Optimization algorithms includes studies of artificial systems based on search behavior of ant colony in the real world and it is used to solve discrete optimization problems (Handl & Meyer, 2002). The Ant Colony Optimization, meta-heuristic algorithms were first proposed by Dorigo, Di Caro and Gambardella in 1999 (Handl, Knowles, & Dorigo, 2003).

Vertical fragmentation algorithm of distributed data based on Ant Colony Optimization clustering technique is presented in this article with the main idea that is based on the smell trace of moving ant colony. The more ants the attribute is selected by, the higher the smell trace concentration is. Accordingly, using the Ant Colony Optimization method in picking (dropping) attributes based on the similarity of the attributes, compared to the preset conditions, to classify the attributes into different clusters such that the similarity of the attributes in the cluster is higher than the similarity of the attributes included in other clusters (Handl & Meyer, 2002).

Ant Colony Optimization based clustering problem uses meta-heuristic clustering style (also ant clustering algorithm) which stems from the basic clustering algorithm in data mining. It is also convenient for us to test results consistent with the results of vertical fragmentation of primitive data under bond energy algorithm (Ma, Schewe, & Kirchberg, 2007) and the improved vertical fragmentation techniques VFC, KO that we have proposed for the results in (Luong, Nguyen, & Le, 2015; Luong, Nguyen, Le, 2015).

The contents of the article are organized as follows: Section 2 presents some related concepts of fragmentation techniques base in clustering under Ant Colony Optimization method. Proposal FAC and VFAC algorithms of vertical fragmentation of distributed data based on Ant Colony Optimization clustering technique is presented in Section 3. Section 4 respectively presents the experimental installation on VFAC and compares and assesses test results using Ant Colony Optimization with other clustering techniques such as k-Means. Section 5 concludes the article (Dorigo & Di Caro, 1999; Handl & Meyer, 2002).

Complete Article List

Search this Journal:
Volume 14: 1 Issue (2024)
Volume 13: 1 Issue (2023)
Volume 12: 4 Issues (2022): 3 Released, 1 Forthcoming
Volume 11: 4 Issues (2021)
Volume 10: 4 Issues (2020)
Volume 9: 4 Issues (2019)
Volume 8: 4 Issues (2018)
Volume 7: 4 Issues (2017)
Volume 6: 4 Issues (2016)
Volume 5: 4 Issues (2015)
Volume 4: 4 Issues (2014)
Volume 3: 4 Issues (2013)
Volume 2: 4 Issues (2012)
Volume 1: 4 Issues (2011)
View Complete Journal Contents Listing