INTERNATIO NAL JO URNAL O F M ULTIDISC IPLINARY EDUC A TIO NAL RESEARC H

ISSN: 2277- 7881; IMPAC T FAC TO R – 6.014; IC V ALUE:5.16; ISI V ALUE:2.286

V O LUME 8, ISSUE 12(6), DEC EMBER 2019

MALL CUSTOMER SEGMENTATION USING CLUSTERING

ALGORITHM

K. Archana Dr.K.G.Saranya
ME Software Engineering Assistant Professor
PSG College of Technology Department of Computer Science
Coimbatore PSG College of Technology
Coimbatore
Abstract
Customer segmentation is a separation of a market into multiple distinct
groups of consumers who share the similar characteristics. Segmentation of
market is an effective way to define and meet customer needs. Unsupervised
Machine Learning technique K-Means Clustering Algorithm is used to perform
Market Basket Analysis. Market Basket Analysis is carried out to predict the
target customers who can be easily converged, among all the customers. In
order to allow the marketing team to plan the strategy to market the new
products to the target customers which are similar to their interests.
Key words: Target Customers, Clusters, Segmentation, Market Basket Analysis
1. Introduction
Management of customer relationship has always played a vital role to
provide business intelligence to organizations to build, manage and develop
valuable long term customer relationships.
The importance of treating customers as an organizations main asset is
increasing in value in present day and era. Organizations are investing
rapidly in the development of customer acquisition, maintenance and
development strategies.
The business intelligence has a vital role to play in allowing companies
to use technical expertise to gain better customer knowledge and Programs
for outreach.
By using clustering techniques, customers with similar means are
clustered together. Customer segmentation helps the marketing team to
recognize and expose different customer segments that think differently and
follow different purchasing strategies.
Customer segmentation helps in figuring out the customers who vary in
terms of preferences, expectations, desires and attributes. The main purpose
of performing customer segmentation is to group consumers with similar

www.ijmer.in 94
INTERNATIO NAL JO URNAL O F M ULTIDISC IPLINARY EDUC A TIO NAL RESEARC H
ISSN: 2277- 7881; IMPAC T FAC TO R – 6.014; IC V ALUE:5.16; ISI V ALUE:2.286
V O LUME 8, ISSUE 12(6), DEC EMBER 2019

interest so that the marketing team can converge in an effective marketing

plan.
Clustering is an iterative process of knowledge discovery from vast
amounts of raw and unorganized data. Clustering is a type of exploratory
data mining that is used in many applications, such as machine learning,
classification and pattern recognition.
2. Related Work
Sanjana et al [1], presented various clustering algorithms taking into
account the characteristics of Big Data such as size, noise, dimensionality,
algorithm calculations, cluster shape and presented a brief overview of the
various clustering algorithms grouped under partitioning, hierarchical,
density, grid-based and model-based algorithms.
Azarnoush Ansari et al [2], Merged clustering of fuzzy c-means and genetic
algorithms to cluste steel industry customers, By using the LRFM variables
(length, recency, frequency, monetary value) system, customers were divided
into two clusters.
Pedro Quelhas Brito et al [3], Investigated different approaches to data
mining for customer segmentation, clustering and sub-group discovery. The
models obtained created six market segments and 49 rules that in a highly
customized fashion manufacturer (e-tailor) allowed a better
understanding of consumer preferences. The complexity of these clustering
DM techniques and their limitations will lead to further methodological
refinement.
Shreya Tripathi et al [4], explored the necessity of segmentation of the
customers using clustering algorithms as the core functionality of CRM.
The mostly used K-Means and Hierarchical Clustering were studied and the
advantages and disadvantages of these techniques were highlighted. At last
the idea of creating a hybrid approach is addressed by integrating the above
two strategies with the potential to surpass the individual designs.
Kishana R. Kashwan et al [5], Presented a comprehensive report on k-
means clustering methodology and SPSS Tool to develop a real-time and
interactive framework for a specific supermarket to forecast sales in various
annual seasonal cycles. The prototype developed was a smart device that
collected inputs at the end of the day's business directly from sales data
records and automatically modified segmentation statistics. An ANOVA
analysis was also performed to test the cluster's stability. The actual sales

www.ijmer.in 95
INTERNATIO NAL JO URNAL O F M ULTIDISC IPLINARY EDUC A TIO NAL RESEARC H
ISSN: 2277- 7881; IMPAC T FAC TO R – 6.014; IC V ALUE:5.16; ISI V ALUE:2.286
V O LUME 8, ISSUE 12(6), DEC EMBER 2019

figures of the day to day are contrasted with the model's expected statistics.
The findings were positive and demonstrated a high degree of precision.
3. Methodology
Clustering
Clustering is one of the most common methods used in exploring data
to obtain an clear understanding of the data structure. It can be characterized
as the task of finding the subgroups in the complete dataset. Similar data is
clustered in the same subgroup. A cluster refers to a collection of
aggregated data points due to some similarities. Clustering is used in Market
basket analysis used to segment the customers based on their behaviors.
K Means Clustering Algorithm
K Means Clustering is the most common and simplest Machine learning
algorithm and it follows an iterative approach which attempts to partition
the dataset into different “k” number of predefined and non overlapping
subgroups where each data point belongs to only one subgroup.
4. Proposed System
It is a web application for segmentation of mall customers which is
useful for the shopkeepers to market the product based on the planned
strategy. The cluster which is generated by the application is stored in the
image format.
Login module:
The marketing team logins the application using the username and
password and each time the marketing team should register with the system
in order to access the details.
Stimulus: Marketing team enters the username and password.
Response: Navigates to the Mall Customer Segmentation Registration page.
Register Module:
The web application provides registration feature for the marketing
team. After successfully logged into the system, each time the marketing
team should register with the system in order to access the details.
Stimulus: Marketing team enters the registration details.
Response: Navigates to the Upload Dataset Module.
MCS - Upload Dataset Module:
The Marketing team is provided with MCS dataset feature in which the
dataset will be imported and it gets navigated to the K-Means execution

www.ijmer.in 96
INTERNATIO NAL JO URNAL O F M ULTIDISC IPLINARY EDUC A TIO NAL RESEARC H
ISSN: 2277- 7881; IMPAC T FAC TO R – 6.014; IC V ALUE:5.16; ISI V ALUE:2.286
V O LUME 8, ISSUE 12(6), DEC EMBER 2019

module. In this module MCS dataset will be imported and after the
display imported successfully message it gets navigated to the K-Means
Algorithm execution module.
Stimulus: Marketing team imports the dataset.
Response: The system provides the message dataset imported successfully
MCS – K-Means Algorithm Module:
The K-Means algorithm will be executed that provides five different
clusters which indicate the customers based on spending score and annual
income.
Stimulus: The marketing team presses the K-Means execution button.
Response: It navigates to the visualization page.
Elbow Method
Elbow method is a tool used for analysing the clusters formed from
our dataset and helps to interpret the appropriate number of optimal clusters
in dataset. From this method the optimal number of clusters for our dataset
is found to be five.

Figure 1: Elbow method – Optimal number of clusters

Visualization Module:
The visualization module provides the results based on the following
clusters. The results are generated in the form of a graph and stored as
image which is retrieved for the marketing team.

www.ijmer.in 97
INTERNATIO NAL JO URNAL O F M ULTIDISC IPLINARY EDUC A TIO NAL RESEARC H
ISSN: 2277- 7881; IMPAC T FAC TO R – 6.014; IC V ALUE:5.16; ISI V ALUE:2.286
V O LUME 8, ISSUE 12(6), DEC EMBER 2019

Figure 2: Customer segmentation result

 Cluster 1 (Red Colour) represents earning high but spending less
 Cluster 2 (Blue Colour) represents average in terms of earning and
spending
 Cluster 3 (Green Colour) represents earning high and also spending
high [Target customers]
 Cluster 4 (cyan Colour) represents earning less but spending more
 Cluster 5 (magenta Colour) represents earning less, spending less
Stimulus: The marketing team presses the K-Means execution button.
Response: The results are generated in the form of a graph and stored as
image. With clusters K=5.
Conclusion and future work
Consumer data is increasing exponentially due to with this large data
magnitude. Such clustering models need to have the ability to effectively
process this huge information.
References
[1]. T. Sajana, C. M. Sheela Rani and K. V. Narayana “A Survey on
Clustering Techniques for Big Data Mining”, Indian Journal of Science and
Technology, Volume 9, Issue 3, Jan 2016.

www.ijmer.in 98
INTERNATIO NAL JO URNAL O F M ULTIDISC IPLINARY EDUC A TIO NAL RESEARC H
ISSN: 2277- 7881; IMPAC T FAC TO R – 6.014; IC V ALUE:5.16; ISI V ALUE:2.286
V O LUME 8, ISSUE 12(6), DEC EMBER 2019

[2]. Azarnoush Ansari, Arash Riasi, “Customer Clustering Using a

Combination of Fuzzy C-Means and Genetic Algorithms”, International
Journal of Business and Management, Volume 11, Issue 7, 2016.
[3]. Pedro Quelhas Brito, Carlos Soares, Sérgio Almeida , Ana Monte,
Michel Byvoet, “Customer segmentation in a large database of an online
customized fashion business”, Robotics and Computer Integrated
Manufacturing ,2015.
[4]. Shreya Tripathi, Aditya Bhardwaj, Poovammal E, “Approaches to
Clustering in Customer Segmentation”, International Journal of Engineering
and Technology, Volume 7, 2018.
[5]. Kishana R.Kashwan, C.M.Velu, “Customer Segmentation Using
Clustering and Data Mining Techniques”, International Journal of
Computer Theory and Engineering, Volume 5, Issue 6, December
2013.

www.ijmer.in 99