In the last decade; distributed database (DDB) has been developed for solving the problem of huge amount of uploaded data
over different network sites. Fragmentation is one of the most important strategies/solutions in order to distribute the organization's
database according to its three type (vertical, horizontal, mixed/hybrid). Furthermore, it reduces the amount of irrelevant data, disk access,
and a single source bottleneck. In this paper, a modified vertical fragmentation strategy that is divided any single relation's attributes into
two or more partitions but by using the advantages of fuzzy logic concept. The modified strategy allows the distributed database designer
to specify the weight (membership function) of all attributes' importance before partitioning it in order to get accurate projection. The
application of this strategy will be led to accurate fragmentation strategy.
Published In:IJCSN Journal Volume 6, Issue 4
Date of Publication : August 2017
Pages : 505-508
Figures :02
Tables : 04
Gamal H. Eladl : Information Systems Department, Mansoura University, Faculty of Computers and Information Sciences,
El-Mansoura, 35516 D.K, Egypt.
DDB, Vertical Fragmentation, Projection, Fuzzy logic, Weight, Membership Function
Distributed database is an important research area because
it needs more effort to utilize the available resources with
the correct placement of data/fragment and the
program/application across a computer network sites. The
fragmentation represents the unit of distribution. The
success of any fragmentation strategy depends on the
accurate selection/choice of appropriate partitions on the
suitable site according to the organization's need. We
explore the vertical fragmentation technique using the
concept of fuzzy logic. As mention above, all the time
fuzzy is better than crisp logic. It deals with numerous
ranges of data (numeric or linguistics). A guideline steps
are introduced to convert any system into a fuzzy-based
system. DDB's designer is responsible for determining
and calculating the importance of any attribute (weight) by
using the application's query log to assign the accurate
value of its membership function. This modified vertical
fragmentation strategy can be applied to any optimized
technique to find the best fragmentation allocation.
[1] S.K. Verma," Fragmentation techniques for distribution
database: A Review", IJICSE, Vol. 3, Issue 2; April
2016.
[2] A. Patel , et. al " A review on Fragmentation
Techniques in Distributed Database", IJMTER, Vol. 01,
Issue 03, September 2014.
[3] R. Basseda, "Fragment Allocation in Distributed
Database Systems", online available at
https://pdfs.semanticscholar.org/00dd/a697ceef89bea2
d4c9fddf738aa54b48bd06.pdf last accessed on May
2017
[4] A.B. Deshmukh, et. al. "Horizontal Fragmentation
Technique in Distributed Database", IJSRP, Vol.2,
Issue 5, May 2015
[5] Suksawat, Bandit. "Diameter Error Prediction Using
Fuzzy Logic for Cast Nylon 6 Turning Operation",
IERI Procedia, 2014.
[6] Huang, Shouren, Yuji Yamakawa, Taku Senoo, and
Masatoshi Ishikawa. "A Pre-compensation Fuzzy
Logic Algorithm Designed for the Dynamic
Compensation Robotic System", International Journal
of Advanced Robotic Systems, 2015.
