An effective adaptive approach for joining data in data

IJRET: International Journal of Research in Engineering and Technology eISSN: 2319-1163 | pISSN: 2321-7308
__________________________________________________________________________________________
Volume: 03 Special Issue: 07 | May-2014, Available @ http://www.ijret.org 257
AN EFFECTIVE ADAPTIVE APPROACH FOR JOINING DATA IN DATA
WAREHOUSE
Sudha.S1
, Manikandan.S2
1
Research Scholar, Adhiparasakthi Engineering College, Chennai, Tamilnadu, India
2
Professor, Computer Applications, RMD Engineering College, Chennai, Tamilnadu, India
Abstract
Formulation of efficient assessment is important for the businesses, because its retrieve lot of details from the data warehouse. In
Data warehouses have materialize as original business information pattern where data store and maintain in concurrent. The
adaptations are requiring in the implementation of Extract Transform Load (ETL) operations. The several methods are included to
joining stream and produce the innovative relation .The previous work was used the adaptive join in data warehouse using ETL
procedure. This approach was common conspire, which create many possible solution. But the drawback of the previous approach is
its not consider exact reproduction. To rise above the question, we are going to present genetic algorithm for joining stream of data .
Several queries and streams are combined in data warehouse the selection of exact grouping of multiple associations are complete via
genetic algorithm. The crossover as well as mutation prefer the paramount consortium of several associations of link for by retrieving
the data and produces the output in data warehouse. The performance of the proposed genetic algorithm used to deliver efficient
highest join data and increase the scalability.
Keywords: join, stream, relation, Genetic algorithm.
-----------------------------------------------------------------------***-----------------------------------------------------------------------
1. INTRODUCTION
Different way we have to store the data in different database.
Genetic Algorithms are powerful search techniques used to
solve many difficult problems. Despite the great successes
achieved in real-world applications, GAs have some
drawbacks. In the genetic algorithm lot of fitness calculations
are necessaryed before an acceptable solution can be found.
Fitness evaluation is not easy in many real-world applications.
There are several situations, in which the fitness evaluation
becomes computationally difficult, so, GAs can be very
demanding in terms of computation and GAs are frequently
used to solve search and optimization problems since they
were first introduced by Holland [1]. They have gained this
prominence by their robustness and simplicity they offer.
Individuals with higher aptitude have more 26 probability to
survive, to reproduce, and to transmit their genetic
characteristics to future generations. GAs can perform
efficient search operations in problem spaces where it is not
easy to understand the environment. Each potential solution in
the search space is considered as an individual (phenotype).
Individuals are represented by using strings that are called
chromosomes. Genes are the atomic parts of chromosomes
and they codify a specific characteristic of a chromosome.
There are several approaches to encode individuals for a
variety of applications. GAs generate a an initial population at
the first phase of the algorithm and then, selection (for mating
and carrying its genes to the next generation, giving higher
probability to "fitter" individuals representing better solutions)
crossover (method for combining genes of two mating
parents), and mutation operations are applied randomly on the
current generation, creating the next generation of solutions
[3]. Individual having the best fitness in the population is the
proposed solution of the problem. For each pair of mating
individuals, the parents' chromosomes are split in two (or
more) parts and genes selected from both parents are
combined to generate a new chromosome which joins the
population. Mutations are also possible where an individual's
randomly selected gene is mutated. Mutations prevent
stagnation and enable a wider exploration of search space that
could not be reached with crossover operators as it is limited
with the available genes in the current population [2]. In order
to keep the population size constant a method is defined for
selecting the individuals that will be copied to the next
generation, and another iteration begins. Termination can be
based on either by production of a fixed number of
generations or the algorithm can terminate when the amount of
improvement in the overall generation quality (e.g. average
fitness values for individuals in a generation) falls below a
predetermined threshold. Also, GA can be terminated at any
given time reporting the best currently available individual as
the discovered solution, thus making GA very suitable for
real-time problems where only a fixed amount of time is
available for optimization.
After an introduction ETL and Genetic Algorithm in Section1,
related literature review of the joining algorithms are given in
Section 2. Section 3 defines the main process models and the
mutation and offspring structure . Evaluation of performance

__________________________________________________________________________________________
and the related results are presented in Section 4. Concluding
clarification and the future works are given in Section 5.
2. RELATED WORKS
Hash-Merge Join (HMJ) (Mokbel et al. 2004), also one from
the series of symmetric joins, is based on push technology and
consists of two phases, hashing and merging. All three
approaches above do not consider the metadata about the
stream. Therefore, they are unable to identify the data which is
no-longer required and by plummeting it the overhead can be
reduced. In addition, the join approaches above focus on
throughput optimization while ignoring the other optimization
goals such as the characteristics of stream data which are
correspondingly important.
The MESHJOIN (Mesh Join) algorithm (Polyzotis et al. 2007)
(Polyzotis et al. 2008) has been presented with the objective to
amortize the slow disk access with as many stream tuples as
possible. To perform the join, the algorithm keeps a number of
hunks of stream in memory at the same time. In each iteration
the algorithm loads a disk partition into memory and attains
the join with all these stream chunks. The algorithm performs
tuning for efficient memory transfer among the join
components, but It is identified in the past some issues around
the access to the disk based relation. Also MESHJOIN cannot
deal with intermittency of the stream competently.
R-MESHJOIN (reduced Mesh Join) (Naeem et al. 2010) is an
enhanced form of MESHJOIN in which one issue related to
suboptimal distribution of memory among the join
components is resolved. However, R-MESHJOIN implements
the same strategy as the MESHJOIN algorithm for retrieving
the disk-based relation.
A partition-based approach (Chakraborty et al. 2009) has been
introduced to deal with intermittency in the stream. It uses a
two-level hash table to attempt to join stream tuples as soon as
they arrive, and uses a partition-based waiting area for the
other stream tuples. The authors do not provide a cost model
for their approach. In addition, the algorithm needs a clustered
index or an equivalent sorting on the join attribute and it does
not prevent starvation of stream tuples.
One recent algorithm, HYBRIDJOIN (Hybrid Join) (Naeem et
al. 2011) address the issue of retrieving the disk-based
relation. An active strategy to access the disk-based relation is
introduced in HYBRIDJOIN.
Another advantage of HYBRIDJOIN is that it can deal with
burst streams, which is a curb of both MESHJOIN and R-
MESHJOIN. However, if it is consider long-tail distributions,
it is find that the algorithm can be improved further.
3. SYSTEM MODEL
The proposed work used Genetic Algorithm is designed here
to perform the multi-join operation efficiently in active data
warehouse. Normally, in active data warehouse, the
information should keep up-to-date with recent values in the
database. So, to add the recent values with active data
warehouse, the genetic algorithm is used here by performing
the highest joining operation in the source streams. The
architecture diagram of the proposed work to perform the
highest joining operation using GA is shown in Fig 1
Fig. 1 System Architecture Diagram of highest joining
operation using Genetic Algorithm
In this system model we can insert the current data to active
data warehouse, the GA use to verify the exact data and add it
in to the warehouse. so this model explain task to execute the
highly join using GA. The main component of the figure .1.
are crossover estimation and mutation, while the source stream
S and relation Rare the input. We first select the relation R
based on the source stream and perform the highly join
operation using Genetic Algorithm process.
Initial relations are loaded in the relation table, the quality of
relations are evaluated based on the fitness .so the initial
relation store in the relation table.
4. PERFORMANCE EVALUATION
The performance of GA based highly join operation in data
warehouse are considered the following ,Data retrieval ,
Efficiency of join operation , Scalability
Table 1
No of Source Max Process Rate
95 1000
100 2800
189 4000
300 4450
670 5600

__________________________________________________________________________________________
Table 1 given details about the number of source arrival in the
method and maximum process rate. Based on this its
fashioned the following presentation.
5. CONCLUSIONS
In this paper, we propose a genetic algorithm based to improve
the performance of our existing join data. This work carry out
highly join operation with relation table in limited
remembrance by using genetic algorithm. An efficient way to
retrieve the data using this GA,GA effectively perform highly
join data by selection ,crossover and mutation.
REFERENCES
[1]. Holland, J.H. Adaptation in Natural and Artificial
Systems.University of Michigan Press, 1975, Ann Arbor, MI,
USA.
[2]. Sevinc, E., and Cosar, A. An Evolutionary Genetic
Algorithm for Optimization of Distributed Database Queries.
The Computer Journal, vol.54, issue: 5, 2011, 717-725.
[3]. Swami, A., and Gupta, A. Optimization of large join
queries.Proceedings of ACM SIGMOD Conf. on Management
of Data, Chicago, Ill, May,1988, 8–17.
[4]. S.Sudha and S.Manikandan,” ADAPTIVE APPROACH
FOR JOINING AND SUBMISSIVE VIEW OF DATA IN
DATA WAREHOUSE USING ETL”, Indian Journal of
Computer Science and Engineering (IJCSE), Vol. 4 No.3 Jun-
Jul 2013 Pp.250-252
[5]. Naeem, M. A., Dobbie, G. & Weber, G. (2011),
‘HYBRIDJOIN for Near-real-time Data Warehousing’,
International Journal of Data Warehousing and Mining
(IJDWM), IGI Global.
[6]. Naeem, M. A., Dobbie, G. & Weber, G. (2011), X-
HYBRIDJOIN for Near-real-time Data Warehousing, in
‘Proceedings of 28th British National Conference on
Databases (BNCOD ’11)’, Springer, Manchester, UK, pp. 33–
47.
BIOGRAPHIES
Sudha.S received the Master Degree in Computer
Applications from the Bharathidhasan University, Trichy,
India in the year 2000, and the M.Phil degree from the same
University in the year 2003.
She is teaching profession for the past 13 years. Previously she
worked as a Lecturer in the Department of M.S and B.Teh(IT).
She is currently an Assistant Professor; Adhiparasakthi
Engineering College affiliated to Anna University, Chennai,
India from 2003 to till. She is currently perusing here PhD
degree in Department Computer Science, Anna University,
Chennai, India. She has published 2 papers in refereed
journals and 5 papers national and international conference
proceedings.
Dr. S. Manikandan received B.Sc. degree in Mathematics
from Madurai Kamaraj University, Madurai in 1996 and
M.C.A. degree from Bharathidasan University, Tiruchirapalli,
in 1999. He received his M.Phil. degree in Computer Science
from Manonmaniam Sundaranar University, Tirunelveli, in
2003 and Ph.D degree from Anna University Chennai in the
year 2009.
He is teaching profession for the past 11 years. Previously he
worked as an Assistant Professor, PSNA Engineering college,
Dindigul. He is currently Director in the Department of
Computer Applications RMD engineering College, Chennai
.He has published six research articles in International and
National journals and presented twenty papers in refereed
national & International Conferences.
0
1000
2000
3000
4000
5000
6000
7000
0 500 1000
processrate
No of source
Max
Process
Rate
Linear (Max
Process
Rate)

An effective adaptive approach for joining data in data

More Related Content

What's hot (16)

Viewers also liked (20)

Similar to An effective adaptive approach for joining data in data (20)

More from eSAT Publishing House (20)

Recently uploaded (20)

An effective adaptive approach for joining data in data