Association Rule Mining (ARM) is now playing a very important role for identifying market strategies and to find out what are the items purchased consistently by various customers. ARM is a data mining function that discovers the probability of the co-occurrence of items in a collection. It is mainly used for focusing on the purchase of the items, to improve the sales strategy consistently with the help of sales transaction data and to maintain the inventory level positively. In this paper, the main objective is on the survey of association rule mining for market basket analysis. This paper provides vast detail about association rule mining which will be useful to minimize complexities of market basket analysis.

1. Integrated Intelligent Research (IIR) International Journal of Business Intelligents
Volume: 05 Issue: 01 June 2016 Page No.90-92
ISSN: 2278-2400
90
An Evaluation of Association Rule Mining
for Market Basket Analysis
S.Sivakumar
Assistant Professor, Dept of MCA,SRM Arts and Science College,Chennai
Abstract-Association Rule Mining (ARM) is now playing a
very important role for identifying market strategies and to find
out what are the items purchased consistently by various
customers. ARM is a data mining function that discovers the
probability of the co-occurrence of items in a collection. It is
mainly used for focusing on the purchase of the items, to
improve the sales strategy consistently with the help of sales
transaction data and to maintain the inventory level positively.
In this paper, the main objective is on the survey of association
rule mining for market basket analysis. This paper provides
vast detail about association rule mining which will be useful
to minimize complexities of market basket analysis.
Keywords-Data Mining, Association Rule Mining, Market
Basket Analysis, The Apriori algorithm, FP Growth algorithm.
I. INTRODUCTION
Today, the business organizations are eager to stand out their
customers and provide more services and experiences. To be
current competitive and stay ahead of line, business
organizations are mostly closing on customer data collection
for insight into trends and customer behavior. However, the
information being collected from various sources is developing
into large quantities of data at a rapid pace. Generating insight
from high velocity and volume requires business organizations
to have a technology landscape that can scale and plays at the
speed of business demands [7,9].Data mining aims to discover
hidden information from massive sets of data [5, 6, 13, 14] and
involves various methods of the intersection of artificial
intelligence, machine learning, statistics and database systems.
Information which is used to increase revenue, cuts costs or
both. Data Mining is the analysis step of the “knowledge
discovery in databases” process. It uses previous statistical data
analysis to derive patterns and trends that exist in data nd to
predict future behavior. Data mining techniques are widely
used in various applications like Financial Data Analysis,
Retail industry, Telecommunication Industry, Biological Data
Analysis, Intrusion detection, Data warehouses, Visualization
and Domain specific knowledge, etc. Association rule mining
[ARM] is an important application of data mining. ARM is a
method for discovering interesting relations between data in
large databases. Most of the business organizations are
concerned in association rule mining from their databases, due
to the increase of large collection of data items in databases.
The chapter 2 presents Association rule mining and their
implementation. ARM is also called Market Basket Analysis
(MBA). This mainly concentrates the buying habits of a
customer by identifying association between the different
items the customer place in their shopping baskets.Section 1
gives the introduction of Association Rule Mining and their
importance, whereas section 2 presents an overview of Market
Basket Analysis and their necessity. Section 3 depicts the
literature survey of MBA. Section 4 covers various
Associations rule mining algorithms and finally conclusion are
depicted by section 5.
II. ASSOCIATION RULE MINING
Association rule mining is used for analyzing and predicting
customer behavior from a large database. It plays an important
role in shopping basket data analysis, product clustering,
catalog design and store layout. This chapter discusses the
importance of Association Rule and their implementation. An
association rule [10] is defined as an implication of the form
x=>y, where x is called antecedent and y is called consequent.
X and Y are set of items and the rule represents if item x are
bought, customer are buy y.
The table 1 shows database with 8 transactions and 4 items
from the supermarket domain.
Table 1: Transaction details
Transacti
on ID
MIL
K
BREA
D
BUTTE
R
BEE
R
1 Yes Yes Yes No
2 No No Yes No
3 No No No Yes
4 Yes Yes Yes No
5 Yes No Yes Yes
6 No Yes No No
7 No No No Yes
8 Yes No No No
From the table 1, Where yes represents the presence of item
and no represents the absence of an item in that transaction. An
association rule is defined as follows {butter,bread}=>{milk}
which means customer would definitely purchased milk if they
had purchased bread and butter.Support and Confidence are
the measurement of the ARM. The ARM support explains the

2. Integrated Intelligent Research (IIR) International Journal of Business Intelligents
Volume: 05 Issue: 01 June 2016 Page No.90-92
ISSN: 2278-2400
91
number of times the customer purchases the milk products
whereas confidence defines as the accuracy of the rule. Based
on table 1, itemset {milk,bread} has a support of
2/6=0.33(33%) whereas the association rule {butter,bread} =>
{milk} has a confidence of 0.3/0.3=1.0(100%). Support is used
for a minimum threshold is supplied to find all frequent items
in a database. Confidence gives a minimum confidence
constraint is applied to those frequent item sets in order to form
rules.
III. LITERATURE SURVEY
Tmka [1] proposed data mining methods for market basket
analysis. This author takes six sigma technique for market
basket analysis implementation. Six sigma techniques provides
various statistics techniques of define(D),
measure(M),analysis(A),improve(I) and control(C) for
improving the results of MBA and modifying the process of
sigma performance level. With the representation of products
dependence, web plot is used.This six sigma method might be
used to target special offers which indirectly improves the
performance level since this can be useful in spending money
which targets a specific customer group. Vo [8] has presented a
model called mining traditional association rules using
frequent itemsets lattice. Various methods have to be proposed
for mining frequent itemsets. With the development of an
application of lattice in mining rule, the time will be decreased
gradually for mining rules. Construction of frequent itemsets
lattice and mining association rules from lattice are the stages
of mining frequent itemsets. With the execution of association
rule mining in lattice, the parent-child relationship is applied.
The result examines that more efficient that the straight mining
from frequent itemsets. In future, the research can propose a
method for considering lattice in mining non redundant rules to
reduce the number of rules by rejecting redundant rules and
reducing the time of mining rules.
Pei [2] has suggested a model of mining association rules
based on Apriori algorithm and application. Mining association
rule has acted a significant subject of data mining research.
This paper also takes mining association rule algorithm and
mining frequent itemsets algorithm on improved measurement
of system. The measure of ARM’s support and confidence and
interestingness useful rules and losing useful rules.Barthik [4]
presents a paper association based classification for relational
data model and its use in web mining. Classification ARM is a
technique for providing better accuracy and human
understandable classification scheme. This paper is to propose
the modification of the basic classification with help of
classification method and used for extracting data from web
pages. This work can be extended the method with the
processing text contained in visual element, which leads to
improved quality of classification.Wang [3] suggested a model
called a novel rule weighting approach in classification
association rule mining (CARM). CARM [12] is a new
classification technique. It creates association rule mining
based classifier with the help of classification association rule
(CAR). CARM algorithm is not considered practically, a
similar set of CAR’s from data and a classifier is given as an
ordered list of cars, established on a selected rule ordering
strategy. This paper introduces a new technique rule weighing
schema, called class-item score rule weighing for providing the
accuracy of classification.
IV. ALGORITHMS FOR ARM
4.1 The Apriori algorithm
This algorithm [11] finds the frequent individual items from
the database and extend them to huge sets. It uses a bottom-up
approach, where recurrent subsets are expanded one item at a
time and group of candidates are examined against the data.
This algorithm will terminate when there is no further
successful extensions are found. There are many difficulties by
applying Apriori algorithm against large databases which scans
the database many times and to take time complexities are high
from frequent item in the larger set length of the
circumstances.
4.2 FP(Frequent Pattern) growth algorithm
This algorithm [11] takes a prefix-tree data structure for storing
compressed and crucial information of frequent items from the
database. It repeatedly creates conditional parameter and from
the FP tree structure and takes them to produce the complete
set of frequent patterns. The divide and conquer strategy are
here applied for decomposing data items from database and the
mining task. It applies frequent pattern method to evade the
costly process of candidate generation and testing opposed to
Apriori algorithm.
V. CONCLUSION
Association rule mining plays a very important role in Data
Mining Techniques. From the collection of large amount of
information from databases, Physical analysis is risky and
difficult. Accuracy and efficiency are important task in such
situations. So it is necessary to apply association rule mining.
This paper covers various association rule mining algorithms
which explains various advantages and disadvantages thus
gives us how these algorithms are used for market basket
analysis. From the Association rule mining algorithms, the
Apriori algorithm is better than any other algorithm for
producing the results from the databases and has some
difficulties. Due to the complexities of such algorithms, there
is a need to improve association rule algorithms for providing
better accuracy and time complexities.
REFERENCES

3. Integrated Intelligent Research (IIR) International Journal of Business Intelligents
Volume: 05 Issue: 01 June 2016 Page No.90-92
ISSN: 2278-2400
92
[1] Tmka, A., “Market basket analysis with data mining methods”,
International conference on Networking and Information technology
(ICNIT), pp. 446-450,2010.
[2] Pei-ji Wang., Lin Shi,Jin-niu Bai and Yu-lin Zhao., “Mining
Association Rules based on APriori algorithm and Application”,
International Forum on Computer Science-Technology and
Applications, Vol. 1,pp,141-143,2009.
[3] Y.J. Wang., Qin Xin and F. Coenen., “A Novel Rule Weighting
Approach in Classification Association Rule Mining”, ICDM
Workshops 2007, Seventh IEEE international conference on Data
Mining Workshops, pp. 271-276,2007.
[4] Barthik.V.,”Association based classification for Relational Data and
its Use in Web Mining”, CIDM ’09, IEEE Symbosium on
Computational intelligence and Data Mining, pp. 252-258,2009.
[5] Jiawei Han and Micheline Kamber,Data Mining Concepts and
Techniques,Morgan Kaufmann Publishers, 2006.
[6] Berry .,M,J.A and Linoff,G, Data Mining techniques for Marketing ,
Sales and Customer Support,USA: John Wiley and Sons 1997.
[7] R.Agarwal., T.Imielinski and A. Swami., “Mining association rules
between sets of items in large databases”, in Proceedings of the 1993
ACM SIGMOD International Conference on Management of Data,
Washington, D.C., 1993, pp. 26-28.
[8] Vo., B ,Le,B.,”Mining Traditional Association Rule Using frequent
items lattice”, International Conference on Computers and Industrial
Engineering, pp,1401-1406,2009.
[9] Andre Bergmann., “Data Mining for manufacturing: Preventive
Maintanence, Failure production, and Quality Control”.
[10] R.Agarwal and R.Srikant., “Fast algorithms for Mining Association
Rules”,Proceeding of the 20th VLDB Conference, Santiago, Chile,
1994,pp. 478-499.
[11] Yen-Liang Chen, Kwei Tang, Ren-Jie Shen and Ya-Han Hu, “Market
Basket analysis in a multiple store environment”, SciVerse
ScienceDirect, Volume 40, Issue 2, August 2005, Pages 339-354.
[12] Li.W, Han . J.,Pei. J, “CMAR:Accurate and efficient Classification
based on multiple-class association rule”, Proceedings of the
ICDM’01, pp. 331-335,2003.
[13] Fu. Y, “Data Mining: Tasks, Techniques and Applications”, IEEE
Potentials, Vol. 16, No. 4, 1997.
[14] K.P. Soman et al, “Data Mining Theory and Practice”, PHI Publishers,
New Delhi, India, 2010.