EXECUTION OF ASSOCIATION RULE MINING WITH DATA GRIDS IN WEKA 3.8

Original Research Paper
1 2 1
Kirti Rathi ,Nejib Doss ,Dr.KanwalGarg
1
Research Scholar, M.Tech. (CSE), Department of Computer Science & Applications, Kurukshetra University, Kurukshetra, Haryana (India).
2
Supervisor, Department of Computer Science & Applications, Kurukshetra University, Kurukshetra, Haryana (India).
EXECUTION OF ASSOCIATION RULE MINING WITH DATA GRIDS IN
WEKA 3.8
ABSTRACT
The premise of this paper is to discover frequent patterns by the use of data grids in WEKA3.8 environment. Workload imbalance occurs due to the dynamic nature of
the grid computing hence data grids are used for the creation and validation of data.Association rules are used to extract the useful information from the large database.
In this paper the researcher generate the best rules by using WEKA 3.8 for better performance. WEKA 3.8 is used to accomplish best rules and implementation of
variousalgorithms.
KEYWORDS: GridComputing,AssociationRuleMining,AprioriAlgorithm,WEKA3.8,Visualizationtools.
1.INTRODUCTION
1.1GridComputing
Grid Computing is considered as one of the promising platform for data and
computation intensive applications like data mining. A grid is basically defined
as data grids that are useful for hardware as well as software that provides
dependable, consistent and transparent access for large scale distributed
resources and shared by various multiple domain organizations in order to
provide support for wide range of applications with the quality of service [1].
Grid Computing is basically a form of networking. Grid Computing is a form of
data grids that is used for distributed and large scale cluster computing and it is a
form of network distributed parallel processing. With the help of data grids in
GridComputingitis easytodiscoverorgeneratefrequentitemsets[2].
1.2Association RuleMining
Association Rule Mining is one of the most important technique among the data
mining techniques that is used for finding the interesting correlation, frequent
patterns, associations or structures among the voluminous and transactional
database. Usually Association Rule Mining is widely used in areas like
telecommunication networks, marketing and inventory control etc. By using
Association Rule Mining minimum support and confidence can easily be
definedfromadatabase[3].
Association RuleMininghas basicallytwo problems:
Firstly find those item sets whose occurrence exceeds a already predefined
threshold in database.These item sets are called frequent item sets.This problem
can further divided into two subproblems candidate large itemsets generation
processandfrequentitemsetsgenerationprocess.
Second problem is to generate Association Rules related with those frequent
itemsetswiththeconstraintsofminimalconfidence[4].
2.APRIORIALGORITHM
AprioriAlgorithm is used to operate on large database that contains transactions
that is collection of items. Each transaction is a set of items or itemset. A
threshold is already defined in apriori algorithm. Itemsets in the transaction must
besubsetofthatthresholdvalue[5].
Let D be a dataset. If p and q are itemset such that q is a subset of p then support of
q is greater or equal to the support of p.All for the need of itemsets to be frequent
all its subsets must be in frequent manner.Algorithm makes multiple passes over
the data. In the every pass item set should be frequent and discovers frequent
itemset of next bigger size. In the first pass it generates all the frequent-1 itemset.
After this by using a second pass with the combination of first pass it generates
frequent-2 itemset by determining their supports. Similarly second pass is used
to generate frequent-3 item set with their support. The candidate generation in
the apriori algorithm reached to the mth passes or an m itemset is considered as
candidateonlyifall(m-1)itemsetscontainedinit[6].
ThetwomainstepsofApriorialgorithmare:
Ÿ The Join step:To find Lk a candidate k-itemsets is generated by joining Lk-1
withitself.Thissetof candidateisdenotedbyCk.
Ÿ The prune step: In the prune step, delete all the itemsets related to the Ck,
where(k-1) subsetofcis notinLk-1.
Ÿ Pseudo codeofAprioriAlgorithm:
Apriori(T, minsupport)
{//Tisthedatasetandminsupportistheminimumsupport
L1={Large–itemsetorfrequentitemset};
For (k=2;Lk-1=null;k++)
{
Ck=candidategeneratedfromLk-1
//Ck=cartesianproductLk-1*Lk-1
//Eliminatek-1 itemsetthatis notfrequent
{
IncrementallthecandidateitemsetsinCk thatareinT
Lk=candidatesinCkwithminsupport.
3.WEKA3.8
Weka 3.8 is a data mining system that is discovered by the University of Waikato
in Newzealand for the implementation of data mining algorithms.Weka 3.8 is an
open source data mining toolkit that is used for performing data mining tasks. It
is a collection of machine learning and visualization tools for easy access with
graphical user interface (GUI) for data mining algorithms. The version of Weka
works with the modeling method and implemented in other programming
languages[7].
Usually Weka 3.8 is an open source data mining toolkit and works in the form of
data grids. Weka 3.8 is based on the various data mining phases: data
preprocessing, classification, association Rules, Regression, Clustering and
Visualization. In this data is available in the form of a file or relation in which
each data point is considered with a fixed number of attributes (numeric or
nominal).
Executionusing Data GridinWeka:
Weka 3.8 is used for the implementation of data mining algorithms. InWeka data
gridsareusedforthecreationandvalidationof variousalgorithms.
Grid is basically considered as large scale support and even used for high
performance support. Management and Scheduling of resource is a very
complex task in a Grid system. In Grid environment it is very difficult to perform
scheduler performance in a repeatable and controllable manner. For the solution
of this problem this paper presents a Weka 3.8 framework that provides
visualization and Simulation method and used Weka toolkit for supporting
distributeddataminingon Gridenvironment[8].
Weka 3.8 is a platform independent and free available open source tool that is
used for the real world data mining applications. In Weka 3.8 the algorithms are
directly applied to the dataset. By using a dataset in the form of excel file formats
and after that it is converted into .csv format and after that it is again converted
into .arff format that is used in Weka 3.8 to analyze the results by using apriori
algorithm.
4.RESEARCH METHODOLOGY
Research methodology is used for the analysis and interpretation of data in this
process work is implemented in Weka 3.8 by apriori algorithm with the help of
datagrids.
Copyright© 2017, IEASRJ.This open-access article is published under the terms of the Creative CommonsAttribution-NonCommercial 4.0 International License which permits Share (copy and redistribute the material in
anymediumorformat)andAdapt(remix,transform,andbuilduponthematerial)undertheAttribution-NonCommercialterms.
7International Educational Applied Scientific Research Journal (IEASRJ)
Computer Science Volume : 2 ¦ Issue : 5 ¦ May 2017 ¦ e-ISSN : 2456-5040

In this research paper data is obtained from “Open Flight Airport Database”. In
this data includes (airports, train stations and ferry terminals). It contains 5888
airlines. Each entry in the dataset includes information of (Airport ID, Name,
City, Country,Timezone,DST, DepartureTimeandArrivalTime)
The attributes are retrieved fromAirports- extended.dat and are implemented in
Weka3.8by obtainingthebestresults.
5.EXPERIMENTALRESULTS
Experimental results are based on the performance in Weka 3.8. As Airline
dataset is used for obtaining the best rules. In the below figure1 researcher get
results by using a visualization tool for obtaining best rules. Various attributes
are used for a particular dataset entry that will give results based on the Apriori
algorithm.
In this paper researcher is using Apriori algorithm that is used to calculate
Associationruleswithminimumsupport andminimumconfidence.By using the
Apriori algorithm in Weka 3.8 researcher will get the effective results with the
increasingperformanceinappropriatemanner.
In the figure1 given below 25 attributes are used in theAirline dataset entry.Thus
it will show the results which concludes process time decreases and threshold
value increases and obtain better results. In the given results it proved that
association have a minimum support as well as minimum confidence and rules
producedbytheassociationwillalsogeneratebetterperformanceresults.
Figure1: Weka 3.8 GUI (preprocess)
For the frequent patterns Association rules are created by analyzing the data
using minimum support and confidence for a particular relationship. For the
creation of association rulesApriori algorithm is used to perform the operations
inWeka.
Implementation inWeka is basically done in arff format or dataset that is used for
the best Association rules. First of all open the file in the preprocess segment
after that it will show two phases. In the left phase it shows all the attributes with
their names. In the right phase it will show the items of the attributes in the upper
portion and in the lower portion it will show the graphical representation of the
attributes.
Next step is to open the associate tab. In the associate tab selection of algorithm is
done. In this research paper the researcher is usingApriori algorithm for finding
the frequent patterns with the help o Association Rules. In the associate tab the
value of the car must be true because it will mine classAssociation rules instead
of (general Association Rules). Then after pressing the start button it will show
the best Association Rules. In the Associate tab figure2 the results are shown
which gave the information of 10 best Association rules by applying Apriori
algorithm.
Figure2: Associate Tab(Associated output)
6.ANALYSISAND INTERPRETATION
After the configuration set up in the Associate tab the best results are shown in
the figure3 with Minimum support and Minimum confidence. Minimum support
is set to 0.7 and Minimimum confidence is set to 0.9 with the 6 number of cycles
in17instances.
In the figure 3 it shows the Minimum support and Minimum confidence with the
number of cycles performed in Associate tab by using Apriori Algorithm. After
this it will generate the best rules by using this Associate tab and increase the
performance of the data grids or frequent patterns that is created by the
AssociationRules.
By the Scheduling method it will increase the performance and shows best rules
for the interpretation of the results with minimum support and confidence. The
nextstepistogeneratethebestrulescreatebytheAssociatetabintheWeka3.8.
Figure 3: Associator Output
Ten best rules found for each class with the different values of Confidence, Lift
leverage and Conviction. After performing the algorithm the results are the
following:
Figure 4: Best Rules
In the above figure finally best rules are found with the help of visualization tool
fortheminimumsupportandconfidencebygeneratingfrequentpatterns.
7.CONCLUSION
In this research paper the researcher used Weka 3.8 with the help of Apriori
algorithm on a large scale dataset of Airlines in terms of the departure time and
arrivaltimefor increasingtheperformanceinaneffectivemanner.
In this research paper performance in both the cases by using Apriori algorithm
as well as by using Weka 3.8 so Weka 3.8 produce the best Association rules for
that dataset after performing the operations on it. Implementation of the Apriori
algorithm is most compatible and the observed results show the effective use of
thedatasetandproducebestrulesafterperformingtheoperationson it.
REFERENCES
1. Anuradha Sharma , Seema verma,” Survey Report on Load balancing in Grid
Computing environment,” International journal ofAdvanced Research & Studies, Vol
4,No. 2, Jan-March,2015.
2. Frederic Magoules, Thi-MAI-houng Nguyen and Lei Yu, Grid Resource Management
toward virtual and services complaint grid computing, CRC, press Taylor & Francis
Group(2009).
3. M.J Zaki ,” Parallel and Distributed Association Mining, a Survey : IEEE
Concurrency, 7(4),pp-4-25,1999.
4. Kumar V, Karypis G. and Han E,” Scalable Parallel Data Mining for Association
Rules,” IEEE Transactions on data knowledge and engineering, Vol 12, No. 3, pp-337-
8 International Educational Applied Scientific Research Journal (IEASRJ)
Volume : 2 ¦ Issue : 5 ¦ May 2017 ¦ e-ISSN : 2456-5040

9International Educational Applied Scientific Research Journal (IEASRJ)
352,2000.
5. R. Agarwal and R. Srikant,” Fast algorithms for mining Association Rules in Large
Databases,”InternationalConferenceonverylargedatabases,1994.
6. P. Tanna and D.Y. Ghodsara,” UsingApriori with Weka for Frequent Pattern Mining ,”
InternationalJournalofTrendsandTechnology,Vol.12,2014.
7. Michael Hahsler and Sudheer chelluboina,” Visualizing Association Rules in
Hierarchical Groups “, 42nd Symposium on the interface: Statistical, machine
learning,andVisualizationAlgorithms(Interface2011).
8. K.R. Swamy and G. H Babu,” Identification of Frequent Item Search Patterns Using
Apriori Algorithm and WEKA Tool”, International Journal of Innovative Technology
andResearch,Vol.3,No. 5,pp.2401-2403,2015.
Volume : 2 ¦ Issue : 5 ¦ May 2017 ¦ e-ISSN : 2456-5040

EXECUTION OF ASSOCIATION RULE MINING WITH DATA GRIDS IN WEKA 3.8

More Related Content

What's hot (20)

Similar to EXECUTION OF ASSOCIATION RULE MINING WITH DATA GRIDS IN WEKA 3.8 (20)

Recently uploaded (20)

EXECUTION OF ASSOCIATION RULE MINING WITH DATA GRIDS IN WEKA 3.8