APPLYING DATA MINING IN CUSTOMER RELATIONSHIP MANAGEMENT

International Journal of Information Technology, Control and Automation (IJITCA) Vol.2, No.3, July 2012
DOI:10.5121/ijitca.2012.2302 15
APPLYING DATA MINING IN CUSTOMER
RELATIONSHIP MANAGEMENT
Keyvan Vahidy Rodpysh1
, Amir Aghai2
and Meysam Majdi3
1
Department of IT, Gilan University of Applied Sciences Crescent, Rasht, Iran
keyvan Vahidy@yahoo.com
2
Industrial Management, University Ramseur,Iran
aghaei89@gmail.com
3
Department of Industrial Engineering, University Ayandegan Tonkabon, Tonkabon, Iran
ABSTRACT
In this article we are going to define the overall customer relationship management (CRM) and Data
mining, Factors between the techniques and software to "data mining" in "CRM" and the interaction
between two concepts. For this purpose and after that in past studies and reports on issues of "data
mining" and "CRM" took place between them. The effect of "data mining" and extract latent information
from large databases of valuable customer has made their determination, and maintenance in order to
attract customers through its taken a step forward and ultimately achieve profitability and efficiency are
good.
KEYWORDS
CRM, Data mining, Applying data mining in customer relationship management
1. INTRODUCTION
With the development of growing s technology, the infamous Internet services competitive
process by increasing access to customers According to different approaches to management
companies in connection with the customers’ show different companies for survival in market
competition wished more profit in business need to analyze that in the market will be done[8]. In
fact that the future oriented companies were identified and needs of their own customers’
infection more belief and showed their own customers’ infection is considered a value for them.
Always in competitive markets more successful action. The companies via information in
connection with their own customers’ infection models that high–ranking bought from other
rivals. Infamous their products and services with regard to this information.
In general cycle stages attract customers included 4 due under phase [7]:
1) Customer Identification
2) Customer Attraction
3) Customer Retention
4) Customer Development

16
With time professional internet professional services the scope of the customers And the use of
various kinds of infamous don’t differ marketing approach s this method, the collections of a
multitude of data founded. And the use of various kinds of infamous doesn’t differ marketing
approach s this method, the collections of a multitude of data founded. To a high volume of data
to analyze and vital devotes to obtain one of the best methods in this field has used the
prospect.[11]
Given the prospect with the use of the computer algorithm to analyze data in order to provide
useful information, useful to paid company goals. All the other hand, can be part of this
information to control profile and understand their customers in different aspects [6]
In continuing to follow it through to the concepts of management and communication with
customers prospect given the role and use prospect given management in connection with
customers.
2. CRM
Intensive rivalry with the commercial and marketing organizations for the business of the deal,
the most precious thing for organizations that the "customers’”. In line with the same
organizations instead of marketing and management relationship unique relations with customers,
value-added products to add his eyes that the value of more marketing relationship to the
organization. This requires a long term relationship with customers that require new topics to
bring to the customer relationship management.[5]
The concept of "CRM" may be considered two-stage skills at the first stage carefully to customer
specifications and divided among them the famous "80 percent interest One of the organization
only by 20 percent of the number of customers to you. "
In fact transition than the product of the customers’ pivotal and concentration to the needs of
customers than the characteristics of customers In the second round to follow the principles of the
first in the discouraging this stage is not merely organization should not be satisfied with the
existing usage But customers’ need to develop through the pivotal s method and tools "relation
with customer management" with the customers to pay. [ 1 ] Perhaps you think about divisions
with customer management "is merely a technological solution but this is what is wrong [ 3 ] In
fact "relation with customer management" need to move market with swing voters are elected.
This process under the title "life of the customers’" be mentioned that includes attract new
customers, the increase in customers will benefit the customers [2]
Swift and Parvatiyar in 2001 with four major CRM knew that customer identify,
customer attraction customers, Customer Retention, Customer development these four
dimensions can be regarded as closed-loop customer relationship management. [7]
Customer Identification
CRM begins with the identification of customers. This phase includes the crowd that is likely that
our customers are a lot or a profit for the organization. Phase elements identified customers,
including analyze customers after months on the customers. analyze clients include the search of
the most profitable customers through the characteristics of the constitution customers. While in
the category includes customers division of labor groups to customers after months of the smaller
customers, including relatively similar customers in each group. [7]

17
Customer Attraction
In the third phase which is the main discussion with customer relationship management.
Customers’ satisfaction to the expectations that customer satisfaction with perception is essential
to protect customers Jupiter preserve elements include marketing person to person pointing to the
marketing person who analyze with support. The diagnosis of nose contains provisions change in
Jupiter. Loyalty programs include support activities with the long-term relationships with
customers especially analyze customers shunned, ranking credit, the quality of services or
satisfaction from the loyalty programs [7]
Customer Retention
In the third phase, this is the main argument CRM. Satisfaction of customer expectations with
customer perceptions of satisfaction is the basic condition for maintaining the customer Elements
of customer retention and marketing person to person marketing analysis is supported. Detect and
predict changes in customer. Loyalty programs, including activities aimed at supporting long-
term relationships with customers turn away customers, especially analysis, credit rating, service
quality or satisfaction from the loyalty program [7]
Customer Development
This phase includes a plurality of transactions, transaction value and customer profitability is
discovered. the elements of customer development, can be cited Cross Selling, Up Selling,
including analysis of the value of customer life [2] The customer relationship management
technology from the four categories of strategic, operational, analytical, interactive.
Strategic CRM
Strategic CRM, a part of a management system communication with customers. In that it will try
to achieve macro strategy and work in the form of statement that the attitude and mission and has
found. With a look at Jupiter pivot and look at the necessities of the review to [7, 4]
Operational CRM
This type of customer relationship management in all areas that contain a direct relationship with
the customer is established in these dimensions. Sometimes a customer points of contact with
both of these places that can be and truth in connection with the organization of the customers in
parts of this happens. [7]
Collaborative CRM
Interactive systems based on the customer participation, are seeking to use the infrastructure and
proper tools, communication channels more and more suitable to the organization and customers.
In fact, this kind of system s are seeking to increase with Jupiter communication channels with
the organization and increase the quality of the canals during the period of time period of
customer relationship with the organization[4]
Analytical CRM
Analytical CRM "to analyze data and information on the organization dealsSo according to the
results of this analysis, management strategies can be identified [4]. A general tool for analyzing
the function of the customer "data mining" [1] Through several organizations store customer data

18
and analysis results have been able to achieve profitability and stability in the market
competition.
It can be inferred from what we have said that "CRM" is the process and technology. The
organization uses a comprehensive view of CRM, by focusing on maintaining and developing
relationships with the customers and finally within reach understanding of customers' needs.
3. DATA MINING
With time and increasing data volume and expansion of modern data storage size, nature and
management of statistical methods, Data necessary to evaluate the extraction of knowledge from
the show. The best solution to meet this data mining can be cited. We show in Table under
evolutionary data mining techniques to collect data.[14]
Table 1: The evolution of data collection methods
technology Evolution
Disks of computers data collection(1960)
Relational database(ODBC, SQL) access data(1980)
OLAP (Online analytical processing) data warehouse DSS(1990)
Multi-processing computer professional data mining(2000)
Data mining process to extract information and advanced data analysis, and extracting knowledge
from massive amounts of data in a database format, a data warehouseor any information storage
is saved. In fact very similar to the data mining is the extraction of metals from the mountain .The
data warehouse and data mining information that lies in the extract. [13]
Figure 1: The above diagram shows the architecture of data mining
Basis of a data mining process involves five stages [9] as follows:
• a set of training samples should be chosen, are collected and trimming.
• Type of knowledge: knowledge of the expected data mining techniques used will be specified.
• Knowledge base: the transfer of existing knowledge about the process
• data mining, often a hierarchy of concepts
• Evaluation criteria: criteria value of knowledge gained from data mining, extraction time and
knowledge in what has been the representation of key importance and will help data mining
process

19
• Presentation: usually extracted is determined depending on the type of knowledge. In many
cases there is also suitable for representation.
Figure 2: Data Mining Process
Away from previous definitions of what data mining have included topics such large-scale
knowledge extraction, data mining, data analysis and finding patterns and relationships between
data is safe. [1]
The ultimate goal of data mining, decision support systems and organizational development. Data
mining and knowledge to extract useful information from large volumes of data are discussed.
Data mining, pattern search will contain information in existing data. These models and
algorithms can be descriptive to describe the data or aspects are predictable, i.e., of unknown
variables, other variables used to predict values. Descriptive data mining, followed by activities
or actions of ifs in the past. Data mining and predictive look at history, to predict future behaviour
[1]
Data mining activities toward knowledge discovery [9] includes the following steps:
1) Identify the purpose and scope of its application is clear understanding that what, in what will
be done within a field.
2) Select the data analysis and discovery for purposes of determining the
3) preparation of data, including data cleansing
4) adopt the best methods for achieving the goals of data mining
5) The application of data mining algorithms
6) Evaluation and validation results
7) The use of stabilization and consolidation of results and knowledge discovered
8) making decisions based on the knowledge discovered
At this stage of exploration data that might typically preparation. Include a focus on data, data
conversion, recorded and performed works selected subset of the initial selection process begin.
Then, depending on the analysis, this stage of the process of data extraction may each selection
includes simple and straightforward for a regression model to elaborate heuristic analysis with
using a wide range of graphical and statistical methods to identify relevant variables and
determine the nature and complexity of the models. [9]
Nevertheless, the data mining course, usually with lots of research and inquiry report and they are
confused. But in fact none of them does not include data mining. Data mining done by special
equipment, the search operation based on repeated analysis of the data. The following
characteristics of data mining [10] explain:
Selection of
information
Trimming
information
Aggregation
of
information
Knowledge
extraction
The
extracted
knowledge
representatio
ns
Detailed
knowledge

20
• The hypothesis does not require special
• different types of data, data mining tools, not only to numerical
• Data mining algorithms to automatically create relations are
• data mining needs to be correct and true
• The results of data mining is relatively complex and requires a team of professionals to express
their
If this is indeed the performance of data mining techniques that created a data mining system
interpreter and procedures including data collection, data integration and devotion to perform data
mining deals. Data mining all unusual patterns which have deviated from the normal state and
may lead to fraud could be found. Data mining results should be interpreted in different modes
that can later study, show. Finally, models can be obtained by customers who have the possibility
of fraud, to predict. [1]
Data mining and analysis methods with observer, through algorithms such as neural networks ,
classification and decision tree (C & RT), genetic analysis, Shopping, SOM applicable. In
addition, common algorithms, as well as new algorithms for scientific research or commercial
purposes through academic research projects, are produced. The unique characteristics of data
mining can be outlined as follows [3]:
• Not only the analysis phase, but the study design and data collection also affect
Able to identify and respond to questions clearly. The main advantage and the difference is that
instead of providing them with other techniques in the same time grand strategy, accurate
responses are given to researchers.
• Ability to assess the effect of different variables on the dependent variables are provided.
Ability to search for the exact questions and answers with high complexity in the data collected
provide.
• To help managers to evaluate the impact of future scenarios
modelling of multiple options and help with decision making under uncertainty must move to the
selected route. Four basic approaches to data mining [3] has the following:
1) Classification
2) Regression
3) Clustering
4) Association
Classification: The prospects of this approach to predict customer behaviour Database records by
Classification Based on defined criteria Through partner values mapping function composition
characteristics to difference category allow phonetic typing mentioned in the decision tree
approach Bayesian learning to be used for evaluation[1]
Regression: This model enables to obtain specific models the rhythm of the comet in data
mining. And provide management reports that the usual tool of statistical theory in the discovery
sequence is set [1].
Clustering: in these approach a division of large heterogeneous sample is used on Applications
that can be mentioned the following [1]:
• Information Retrieval
• Compression associated with the destruction

21
Association: The purpose of these communication different items in the collection of information
inputs and finally support and confidence through various measures are related [1].
4. Data Mining in CRM
The important role in the process of data mining "CRM" plays. Because on one hand, with data
canters or data warehouse is the interaction and the interaction with the software analysis is
competitive management. Relationship between the manual management competitive analysis
software with data mining software requires the transcription of data models to the data
transmission. The same definition in the software sector, customer data mining and analysis
software for corporate management to ensure that competitive. The model need not have your
entire database. [12] Data mining in customer relationship management process is shown in
figure 3:
Extraction, translation and call processes provide data Pattern discovery Analysis model
improvement
Figure 3: data mining process in CRM
Raw data are collected from various sources and the extraction, translation and management
processes of this type are called data warehouse. In data preparation, data out of the warehouse
and data mining are an appropriate format. [1, 12]
The overall advantages of using "data mining" in "CRM" [15]:
1) data filtering to eliminate duplicate data
2) extraction, data management, analysis and access to utility customers, retain customers models
3) rapid and accurate access to integrated data
4) The use of precision instruments and advanced data analysis and reporting
5) increase customer satisfaction
6) to attract potential customers, retain existing customers and increase market share
In total, over a period of data mining can be used to predict the exact changes. Organizations in
order to fetch data to identify patterns in data mining helps customers .For example, a company
can predict from the data mining toward a specific event in the "customer life cycle" to use. [1]
In fact, data mining techniques can help to extract or detect the presence of lifetime customers in
the United. And create value for the analysis of performance data and behaviour customer.Our
aim in this study, a review article by Ngai[3] analysis is given in Figure 4.
Raw data CRM Data
Warehousin
g
Provided
data
Pattern
discovery
Hypothesis
predicts a
data
analysis

22
Figure 4: the classification of data mining techniques in CRM
As you have seen in the Figure 4 data mining methods include:
Clustering: Clustering of data mining methods is descriptive, Classification of heterogeneous
data into several clusters based on similar characteristics, that purpose of clustering, classification
data is partial to a few groups Data from the different groups may also have different maximum
and existing data in a group are very similar to the clustering with different classification. And
this difference and not being labelled categories is related to the clustering of pre-defined
categories [1]
Classification: Mapping data in predetermined categories, sorting is done in two stages.
Beginning on the last batch of data is identified and it is predicted that a new data belongs to
which category [1]
Association: In the early 90s with the advent of data mining algorithms to extract association
rules from the database into the field formed. Among the activities that will do this, find the
dependencies and relationships between data, finding patterns that exist between the data, And
also making a series of structured items and objects in the database [1]
Regression: Continuous values predicted by the statistical technique called regression. The
regression analysis model that can determine the output variable with multiple input variables to
determine. Regression analysis, a process that determines how the variable associated with one or
more variable Y Х1, Х2, ..., Хn. Response or dependent variable Y is called the output and input
Хi-Y, returned products, called the independent variable. Indeed, regression techniques, a very
common technique for the study of customer satisfaction level is predicted [1]

23
As you have seen in the Figure 4 CRM dimensions include:
Customer identification: In this article, because the customer segmentation, customer analysis
purposes is discussed. The customers have been used segmentation than clustering techniques.
However, for purposes of customer analysis techniques used are classified.
Customer attraction: Discussed in the following general categories based on mainly through
direct marketing and the techniques used are classified
Customer retention: In the following categories such as complaint management, relationship
marketing is the loyalty program. Clustering method and sequence of discovery in the topic of
complaint management and customer loyalty programs within the topic of the classification
method. However, most research in the field of relationship marketing has been paid to the
Forums.
Customer development:These methods mainly because of the exchange lifetime customer and
market basket analysis, it is raised. Articles is mainly used in the clustering of customers through
the methods and rules based forums.
As noted, articles on current research in 2009 has been reviewed by Negai
Table 2: the classification of data mining techniques in CRM
Amount
Data mining
model
CRM dimensions
5
Classification
Customer
identification 6
Clustering
1
Regression
1
Regression
Customer attraction 5
Classification
1
Clustering
7
Clustering
Customer retention 27
Classification
2
Regression
13
Association
1
Classification
Customer
development
2
Clustering
1
Regression
5
Association
As can be seen in Table 2.Volume of activities carried out in the preservation of other aspects of
customer relationship management, especially after the customer has been attracting customers.
And methods used have been classified and most forums.See Figure 5 in most research studies
related to the classification and clustering methods

24
Figure 5: the extent of data mining techniques used in the customer management between 2000 and 2006
In table 3, Data Mining methods have been investigated and their strengths and weaknesses in
customer relationship management
Table3: Strengths and weaknesses of various data mining techniques in CRM
Weaknesses Strengths
Data mining
techniques
• method's performance alone is not
sufficient to predict customer behavior
[1]
• The most widely used method
• Initial assessment of customer data
[1]
Clustering
• due to lack of stability, stability in
some of the methods are simple[18],
yet complex and obscure methods
are stable and properly
functioning[16]
• Diversity of the simplest techniques
used to the most complex
methods[1]
• provide tangible rules[17]
Classification
• Inability to express behavior
patterns hidden in data
• the inability of the behavioral
patterns of behavioral phenomena [16]
• Ease of application performing
model
• very rich literature on the use of
model[19]
Regression
• Total amount of items that do not
frequent [20]
• Ability to discover hidden
relationships among data behavioral
• the ability to sequence the events,
phenomena customer behavior [20]
Association
Rules
5. Conclusion
Organization of the maxim that the customer behaviour over time will change. Organizations
know more about each other if their relationship will evolve and grow better. Generally, the use
of "CRM" software and technologies used in its proper context for the relationship between the
organizations provides customers. Key to the success of the "CRM” having a good strategy in the
management of the "customer life cycle" .Using a variety of technologies and data mining
applications and methods appropriate for the area between the stored data. Years due to
interaction with customers in different time periods and knowledge needed to achieve marketing
success and profitability for the organization .Procedures to be adopted and causes your target
market to potential customers so we can become the ultimate customers. Generally, the use of
data mining in "CRM" is due to good response. In order to customer needs, improving returns on
capital, labor productivity, Quality improvement in end products that can respond quickly to
environmental change
0
1 0
2 0
3 0
4 0
5 0
6 0
7 0
C lu st erin g C la ssific at io n Re gres sio n A sso c iat io n
C o u n t article
C o u n t artic le

25
REFERENCES
[1] Berry ,Gordon S.linoff., “EBook Data Mining Technique for marketing Sales and CRM: Wiley
Publishing”, Inc., Indianapolis, Indiana ,2004
[2] Gneiser,“Value-Based CRM, The Interaction of the Triad of Marketing Financial Management and
IT”,2012
[3] Ngai , Xiu , Chau,“Application of data mining techniques in customer”, Expert Systems with
ApplicationsVol. 36,pp.2592–2602,2009
[4] Olafsson, S., Li, X., & Wu, S, “Operations research and data mining”, European Journal of
Operational Research , Vol.187, pp.1429-1448,2008
[5] Ryding(2010), “The impact of new technologies on customer satisfaction and business to business
customer relationships: Evidence from the soft drinks industry” , Journal of Retailing and Consumer
Services, Vol.17 pp.224–228
[6] Rygielski , Wang, Yen .“ Data mining techniques for customer relationship management “,
Technology in Society, Vol.24 , pp.483–502,2002
[7] Parvatiyar, Sheth,“Customer relationship management: Emerging practice, process, and discipline.
Journal of Economic& Social Research,Vol.1,pp.1-3,2001
[8] Payne,“Customer Relationship Management: A Strategic Perspective, Gabler Verlag ,Vol.3,2009
[9] LuanJing ,”data minig as driven by Knowledge management in higher education”, Public conference
UCSF, 2001.
[10] Morgan Kaufmarm,”Data mining:concepts and Techniques” .Academic Press Publihers,2001
[11] Reza Allahyari Soeini and Keyvan Vahidy Rodpysh,"Evaluations of Data Mining Methods in
Order to Provide the Optimum Method for Customer Churn Prediction: Case Study Insurance
Industry",2012 International Conference on Information and Computer Applications (ICICA
2012)IPCSIT vol. 24 (2012) © (2012) IACSIT Press, Singapore
[12] Reza Allahyari Soeini and Keyvan Vahidy Rodpysh ,"Applying Data Mining to Insurance
Customer Chu rn Management",2012 IACSIT Hong Kong Conferences IPCSIT vol. 30 (2012) ©
(2012) IACSIT Press, Singapore
[13] Dr. Jamal Shahrabi, "Seminar on Data Mining Society of Iran", January 2010
[14] Dr. Jamal Shahrabi,” Ebook: Data Mining", 2005
[15] Samira Malek Mohamadi,” graduate seminar, Application of data mining in customer relationship
management”, University of Science and Technology,2009
[16] Au , Chan."A Novel Evolutionary Data Mining Algorithm with Applications to Churn Predi
ction.",IEEE Transactions on Evolutionary Computation, Vol. 7, No. 6, Dec 2003,pp. 532-545
[17] Guo-en, Wei ,“ Model of Customer Churn Prediction on Support Vector Machine”, Systems
Engineering Theory & Practice,Vol. 28,2008,pp. 71-77
[18] Hadden , Tiwari , R.Roy ,D. Ruta.(2005). ,” Computer assisted customer churn management: State-
of-the-art and future trends”, Computers &Operations Research, Vol. 34,2005,pp.2902-2917
[19] Hwang, Jung , Suh ( 2004) ,”An LTV model and customer segmentation based on customer value: a
case study on the wireless telecommunication industry”,Expert Systems with Applications,
Vol.26,2004, pp.181–188
[20] Chiang , Wang , Lee , Lin, “Goal-oriented sequential pattern for network banking churn analysis”.
Expert Systems with Applications, Vol 25,2003, pp.293–302

APPLYING DATA MINING IN CUSTOMER RELATIONSHIP MANAGEMENT

More Related Content

Similar to APPLYING DATA MINING IN CUSTOMER RELATIONSHIP MANAGEMENT (20)

More from IJITCA Journal (20)

Recently uploaded (20)

APPLYING DATA MINING IN CUSTOMER RELATIONSHIP MANAGEMENT