ji-erp crm

8/3/2019 JI-ERP CRM

1/10

Abstract As Enterprise Resource Planning (ERP)implementation has become more popular and suitable for everybusiness organization, it has become a essential factor for the successof a business. This paper shows the best integration of ERP withCustomer Relationship Management (CRM). Data Mining isoverwhelming the integration in this model by giving support forapplying best algorithm to make the successful result. This model hasthree major parts, outer view-CRM, inner view-ERP and knowledgediscovery view. The CRM collect the customers queries , EPRanalyze and integrate the data and the knowledge discovery gavepredictions and advises for the betterment of an organization. For thepractical implementation of presented model, we use MADAR dataand implemented Apriori Algorithm on it. Then the new rules andpatterns suggested for the organization which helps the organizationfor solving the problem of customers in future correspondence.

Keywords Apriori Algorithm, CRM, ERP.

I. INTRODUCTIONAs an electronic business environment changes more rapidly

under the globalization, even small and medium sizecompanies also change their business. With enterprisesbecoming bigger and bigger, the legacy business systems maynot be flexible enough to adapt this change and thediscordance between business and information systems in theirorganization may occur [4]. Therefore, recently mostcompanies use an Enterprise Resource Planning (ERP) systemfor improving core competency.Enterprise Resource planning (ERP) integrates thefunctionality of all the business departments in an organizationin a single system to carry out the particular needs of thesedifferent departments and share their information very easily.Interaction channels between enterprises and customers have

The revised version submitted on 30th march 2009. This work wassupported by the Research Center of College of Computer and InformationSciences, King Saud University ,Riyadh, Kingdom of Saudi Arabia.

F. A. Author is an Associate Professor in the department of InformationSystems, King Saud University , Riyadh, Kingdom of Saudi Arabia. Mob:00966-504171443, [email protected].

S. B. Author is a Researcher in the Department of Information Systems,King Saud University , Riyadh, Kingdom of Saudi Arabia. Mob:00966-530706625, [email protected].

T. C. Author is a Researcher in the Department of Information Systems,King Saud University , Riyadh, Kingdom of Saudi Arabia. Mob: 00966-543807381, [email protected].

been gradually changed owing to the development information technology. In the highly specialized businenvironment, the interaction model should be re-engineereenhance the quality of customer service [9].ERPs best hope for demonstrating value is as a sort of battering ram for improving the way your company takecustomer order and processes it into an invoice and reve

otherwise known as the order fulfillment process [14]. Thawhy ERP is often referred to as back-office softwaredoesnt handle the up -front selling process (although moERP vendors have developed CRM software or acquired pplay CRM providers that can do this); rather, ERP takecustomer order and provides a software road map automating the different steps along the path to fulfillin[14]. It is now widely accepted that ERP systems providviable alternative to custom application development for standard information management needs and that it is osuperior in terms of quality of the implemented businprocess [2]. Although current common ERP systems provenough parameters and can be adjusted according to indu

characteristics, they is still too complicated to use oeffective under certain conditions [6]. ERP solutions designed to solve the fragmentation of information businesses, and integrate all the information flowing withcompany are designed to solve the fragmentation information in businesses, and integrate all the informaflowing within a company [12].

II. RELATED WORK / BACKGROUNDAccording toVirgil Chichernea and Romanian [8] the basicarchitecture for an ERP system consists of 12 businfunctions utilizing a common manufacturing database, shown in the following Fig.1. Beyond that CRM ofdescribes a strategic or philosophic approach for managcustomers [1]. Hence CRM could be seen from a procoriented, technological, capability-oriented, philosophiand/or strategic perspective [1]. According to ForresResearch, 57% of business firms cannot justify CRinvestments because they cannot measure customprofitability [1]. The core of the ERP system circulates withe company as well as the management information control needs of the entire production process, includreducing inventory, labor, and operation costs, improvbusiness processes to enhance operation efficiency a

DEVELOPING AN INTEGRATED DATA MININGENVIRONMENT IN ERP-CRM MODEL

A CASE STUDY OF MADAR

A. Abdullah S. Al-Mudimigh, B. Farrukh Saleem, C. Zahid Ullah

Department of Information SystemCollege of Computer and Information SciencesKing Saud University, Riyadh

Kingdom of Saudi [email protected], [email protected], [email protected]

INTERNATIONAL JOURNAL OF EDUCATION AND INFORMATION TECHNOLOGIESIssue 2, Volume 3, 2009

135
mailto:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]

8/3/2019 JI-ERP CRM

2/10

improving customer response [6]. However, Markus andRobey (1988) pointed out that although industry specific ERPhas already focused on industry characteristics and includesthe optimal business operation management model, thepromotion of ERP is still significantly related to interactionwith the organization [6]. The necessary broad level of security is determined by the CRM security objectives whichan organization needs to meet[13]. A CRM security strategy

outlines in general terms how an organization will achieve itsCRM security objectives [13].

In order to increase the use of ERP systems it is recommendedto begin with the financial section, the applications invoicing,

cost control, accounting and financial then it should be addedmany functions from Financial, Relation ManagementProduction, Distribution, e- Business and Analyses [8]. CRM(Customer Relationship Management) systems, which providethe framework for analyzing customer profitability andimproving marketing effectiveness, have become anindispensable component in enterprise information systems.Typically, CRM activities include data analysis, campaigndesign, response analysis of customer data[3]. The order of thecustomer is routing automatically to the next department whenone department finishes their work of the customer order andeach department have access to the single database that holdsthe customers new order [10].

III. METHODOLOGY

The ERP-CRM model shown in the Fig.2 , clearly describedthe ERP model to solve the business problems. Whenevercustomer request engender directly forward to the concerndepartment for the assessment and positive response. Afterstatistical analysis and evaluation of the request the answerback to the corresponding customer and this feedback will besaved in the database for the future requirements. This

scenario is most common in each and every organization creating more sophisticated and steadfast environment presented this whole scenario in our model in three diffeviews. In effect after, there is a construction of an easy to and implement surroundings for the upper management their employees for put their stress and expertise within companys limitation and taking their company on top andbehaving with their customer on priority.

The three views are as follows:i. Outer Viewii. Inner View

iii. Knowledge Discovery View

Description of Three Views

A. Outer ViewWhenever a customer contact with the company the custosupport officer receive customers request. In the companysprospectus this department has much importance becausecorrespondence directly with the customer. In our model presented customer relationship management as an outer vCRM is responsible for receiving requests and replying tocustomer directly. These requests includes quericomplaints, suggestions and orders then forward these requto the inner view enterprise resource planning (ERP) throthe query generator. After taking action on the perspectrequest the answer will forward through the CRM The OuterView. And result will also be save in the database knowledge discovery view.

B. Inner ViewThe important part of the model is inner view or ERP viewthis view each department have equal access to a sindatabase that holds the customers data or complaints. In ERPview the customer queries rotating and evaluating by concern department. For example a customer want to purchany product will apply for a product through the customsupport department (outer view) and the request will forward to the sales department (inner view) and sadepartment will check the payment status in the database will forward the same request to the operations departm(inner view) for the delivery of the product and the record be stored in the database to generate new rules and patte(knowledge discovery view) from the experienced data.

C. Knowledge Discovery View

This view is concern with the central database having all kof data saved from outer and inner views. This data cancustomers history, manufacturing status and sales. In thisenvironment we can use several kinds of data mintechniques for discovering the knowledge. The bimplementation is presented in this paper by using AprAlgorithm.

Fig. 1- Architecture of an ERP System


136

8/3/2019 JI-ERP CRM

3/10

IV . CASE STUDY

For the practical implementation of presented model weapplied data mining technique on MADAR data. MADAR isan ERP based organization physically working in King SaudUniversity dealing with all administrative software of theuniversity and also do work for outer projects. The efficientimplementation of presented ERP-CRM model using datamining techniques is applied on MADAR data. For this weused Association mining- Apriori Algorithm for finding newrules and patterns from the experienced data. The descriptionof all phases on our case study are as follows:

A. Apriori AlgorithmApriori is a classic algorithm for learning association ruApriori is designed to operate ondatabasescontaining

transactions (for example, collections of items bought customers, or details of a website frequentation) [7,11]. Acommon in association rule mining, given a set of itemsets instance, sets of retail transactions, each listing individitems purchased), the algorithm attempts to find subsets whare common to at least a minimum number C of the itemsApriori uses a "bottom up" approach, where frequent subare extended one item at a time (a step known as candidgeneration), and groups of candidates are tested against data. The algorithm terminates when no further succesextensions are found [7, 11, 15]. One way to construc

Fig. 2- ERP CRM Model


137
http://en.wikipedia.org/wiki/Databasehttp://en.wikipedia.org/wiki/Database

8/3/2019 JI-ERP CRM

4/10

simpler model computed from data, easier to understand andwith more predictive power is to create a set of simplified rules[5]. Apriori Algorithm is suitable to compute the rules andpatterns and predict for any organization to improve thecustomer satisfaction. We implement Apriori algorithm onMADAR data and generated some rules and patterns forMADAR. The descriptions of these rules are in the followingsteps.

A.1 . Apriori Pseudo Code

Apriori

large 1-itemsets that appear inmore than transactions }

while Generate ( Lk 1)

for transactions

Subset(C k ,t ) for candidates

return [7, 11, 15].

B. Programming Tool Used

For the implementation of Apriori Algorithm to get new ron the MADAR data we used VB technology. Input and oudata files can be in any format e.g databases, Notepad, Exworksheets, etc. But in this case we used Notepad and Eworksheet as input and output files.

B.1.VB Interface

The VB interface is shown in Fig. 3

Fig. 3- VB Interface for Apriori


138

8/3/2019 JI-ERP CRM

5/10

B.2. VB Code

The VB Code is shown in Fig.4.

C. DATA GATHERING / SURVEY We gathered the data by survey. For this we make a survey forMADAR and collect the two months data through aquestionnaire from each and every department. The data isthen organized in a small database.

C.1. Data Transformation

In the Fig.5. we transferred the useful data from the databinto the excel format for describing the data more sufficieand for applying some rules and formulae on surveyed dFurthermore we columned the data in customer query, quedepartment, corresponding department and action by department.

Fig. 4- VB Code for Apriori


139

8/3/2019 JI-ERP CRM

6/10

C.2. Data Enhancement In the Fig.6. we enhanced the data by categorizing the dataaccording to the appropriate action of the department. In whichwe categorized all three actions of the department in three (03)categories. Further we add one more column to assign binary

value (0 or 1) according to the action from the MADcorrespondent. This extra column is the key for implementation of apriori algorithm which is shown in the steps.

Fig. 5- Transformation of Data

Fig. 6- Data Enhancement


140

8/3/2019 JI-ERP CRM

7/10

C.3. Selected Data for Apriori ImplementationIn the Fig. 7. Shows the selection of concerned data after someenhanced features we used as an input file. We select notepadformat as an input file for the Apriori program developed by

using VB Technology. The selected data for input file incluthe Customer Query and Binary Value of action on the qudid by department. Moreover we put 5% minimum supporan input also for the generation of frequent itemsets frominput file.

C.4. Output DataThe Fig.8. Shown the output file with 5% minimum support.We applied apriori algorithm for the generation of frequentitemsets. These Itemset generation is based on the customerquery, corresponding binary value and minimum support(5%).In consequence we find out that how much queries has beendealt on time, and how much queries were delayed in the last

two month. In the next step we generated and suggested srules from the frequent itemset generations (output file) to tMADAR organization for handling with the customer in mefficient way and for the betterment of the organizationfuture. This was a sample implementation of apriori algorion the CRM-ERP based model. We can find different kindfrequent itemsets generation in the same manner by increaor decreasing the minimum support.

Fig. 7- Selected Data for Apriori


141

8/3/2019 JI-ERP CRM

8/10

C.5. Rules GenerationFig.9. Shown the selected and rejected rules table. After rulesgeneration we found several rules which proposed andsuggested to the MADAR that there are some basic problemand queries are coming from the several department within thelast two months records. The rules is showing the other bigproblems which are actually because of some basic problems.After the implantation of these rules we can decrease number

of queries and several kind of queries will automaticallyresolved. i.e. website domain problem online courseregistration, login problem, mail server problem. For examif the query raised from the website controller then asometimes the query will also be generated from other onrelated department, i.e. online course registration. Wconclude that the basic query was website domain problemthe MADAR will overcome this problem immediately tother queries can be stop.

Fig. 8- Output Data


142

8/3/2019 JI-ERP CRM

9/10

V. FUTUREWORK The model will be enhancing with more information and moredata mining technique will be applying and new rules will begenerated in future for the enhancement of an organization.New ERP tools will be used to modify the existing work andmake ease for the customers to access the organizationsfacilities with out any hesitation.

VI. CONCLUSION In todays technologies the customer have a lot of difficultiesto access the organizations facilities. The customer haveproblem in contacting the organization. The model presentedin this paper will solve these problems all the customerscomplaints will be recording in the central database and willbe process according to the customer need. The customer caneasily contact the organization and can purchase theorganization products very easily. The CRM (outer view) willcollect the information about the products and the queries willbe forwarded to the inner view (ERP) to act upon thesequeries. The knowledge discovery view generates new rulesand patterns for the betterment of an organization for futurecorrespondence to improve the growth of the customers for anorganization.

VII. ACKNOWLEDGEMENT

Our cordially thanks goes to Rector and Vice Rector (KETT)of King Saud University for their financial support. Ourspecial thanks goes to the Dean and Vice Dean College of Computer and Information Sciences King Saud Universitywho encouraged us in the field of research. Also thanks to ourchairman Abdullah S. Alghamdi for each and every thingprovided us for this work. Thanks to Maqsood Mahmood fortheir guidance.

REFERENCES[1]. Sonja Grabner-Kraeuter, Gernot Moedritscher, Martin Waigunyc, WeMussnigb, Performance Monitoring of CRM Initiatives, in Proceedings:IEEE conference on System Sciences, 2007.

[2]. Philippe Dugerdil, Gil Gaillard, Model -Driven ERP Implementation InProceedings: ICEIS, 8th International Conference on Enterprise InformatiSystems, May, 2006, Paphos, Cyprus.[3] Han-joon, Kim TaeHee, Lee Sang-goo, Lee Jonghun Chun, "AutomData Warehousing for Rule-based CRM Systems", 14th AustralaDatabase Conference (ADC-2003), Adelaide, Australia, 2003.

[4]. Jae-won Park, Nam-Yong Lee, A Conceptual Model of ERP for Small

and Medium-Size Companies Based on UML, IJCSNS International Journalof Computer Science and Network Security, VOL.6 No.5A, May 2006.

[5]. Mar__a C. FERN_ANDEZ_, Ernestina MENASALVAS_, MARB_AN_Jos_e M. PE~NA__, Socorro MILL_AN, MINIMAL DECISION RULESBASED ONTHE APRIORI ALGORITHM y, Int. J. Appl. Math. Comput. Sc.i, Vol.11,No.3, 691704, 2001.

[6]. Wen-Hsiung Wu, Chin-Fu Ho, Hsin-Pin Fu, Tien-Hsiang Chang, SMESIMPLEMENTING AN INDUSTRY SPECIFIC ERP MODEL USINGCASE STUDY APPROACH, Journal of the Chinese Institute of IndustrialEngineers, Vol. 23, No. 5, 2006, pp. 423-434.[7].Apriori Algorithm, http://en.wikipedia.Org /wiki/Apriori algori#Algorithm,

Accessed Date: 12th April, 2009.[8]. Virgil Chichernea, Romanian, THE USE OF THE ERP -CRM-CIMSYSTEMS WITHIN THE MASTERS DEGREE PROGRAMMES InProceedings: Annales Universitatis Apulensis Series Oeconomica, Nr.2007, Volume 2.

[9]. Chang-Shing lee, Yi-Chuan Wang, Wei-Ming liu, Yi-Chen lin, CrmOntology Based On CMMI Project Planning For Business applications, inProceedings: IEEE conference on machine learning and cybernetics, HKong, August, 2007.

Fig. 9- Rules Generation


143

8/3/2019 JI-ERP CRM

10/10

[10]. Darwin Publications, "Executive Guides, Enterprise Resource Planning- ERP", http://www.netessence.com.cy/downloads/erp.pdf , Accessed date:March 21, 2009.[11]. Agrawal R, Imielinski T, Swami AN. "Mining Association Rulesbetween Sets of Items in Large Databases." SIGMOD. June 1993, 22(2):207-16.

[12]. Gail Corbitt, Marinos Themistocleous, Zahir Irani, ERP/EAI SystemIssues and Answers: A Research Journey, in proceeding: IEEE conference onsystem sciences, 2005.

[13]. Mahdi Seify, New Method for Risk Management in CR M SecurityManagement, in proceedings: IEEE conference on Information Technology,2006.

[14]. Thomas Wailgum, "Enterprise Resource Planning: Definition andSolutions", CIOhttp://www.cio.com/article/40323/ERP_definition_and_solutions, Accesseddate: March 11, 2009.

[15] S. Kotsiantis, D. Kanellopoulos, Association Rules Mining: A RecentOverview, GESTS International Transactions on Computer Science andEngineering, Vol.32 (1), 2006, pp. 71-82


144

ji-erp crm

Documents