ID3 Algorithm to Identify Customer Loyalty Factor at Semarang Ceramics Company
Tesis
Oleh: Isadora Nugroho NIM: 972011006
Program Studi Magister Sistem Informasi Fakultas Teknologi Informasi Universitas Kristen Satya Wacana Salatiga Juni 2013
ID3 Algorithm to Identify Customer Loyalty Factor at Semarang Ceramics Company
Tesis
Oleh: Isadora Nugroho NIM: 972011006
Program Studi Magister Sistem Informasi Fakultas Teknologi Informasi Universitas Kristen Satya Wacana Salatiga Juni 2013
i
ii
Pernyataan Tesis berikut ini : Judul
Pembimbing
: ID3 Algorithm to Identify Customer Loyalty Factor at Semarang Ceramics Company
: 1. Prof. Ir. Danny Manongga, M.Sc., Ph.D 2. Dr. Ir. Wiranto H. Utomo, M.Kom.
adalah benar hasil karya saya : Nama
:
Isadora Nugroho
NIM
:
972011006
Saya menyatakan tidak mengambil sebagian atau seluruhnya dari hasil karya orang lain kecuali sebagaimana yang tertulis pada daftar pustaka.
Pernyataan ini dibuat dengan sebenarnya sesuai dengan ketentuan yang berlaku dalam penulisan karya ilmiah.
Salatiga, Juni 2013
(Isadora Nugroho)
iii
Prakata Puji syukur ke hadirat Tuhan Yesus Kristus atas berkat, rahmat, serta bimbingan-Nya sehingga penulis dapat menyelesaikan Thesis yang berjudul “ID3 Algorithm to Identify Customer Loyalty Factor at Semarang Ceramics Company”. Laporan thesis ini disusun guna memenuhi persyaratan untuk memperoleh gelar Master of Computer Science (M.Cs.) pada Program Studi Magister Sistem Informasi Fakultas Teknologi Informasi Universitas Kristen Satya Wacana Salatiga. Dalam menyelesaikan thesis ini, penulis mendapat bantuan dan dukungan dari berbagai pihak, baik secara langsung maupun tidak langsung. Oleh karena itu pada kesempatan ini penulis ingin mengucapkan terima kasih kepada: 1.
Tuhan Yesus Kristus yang telah memberi kesehatan kepada penulis sehingga dapat dengan lancar mengerjakan thesis ini.
2.
Papi, Mami, Koko (sigit, eddy, santoso), Cece (olif, diana). Terima kasih atas segala bantuan baik moril maupun materiil, dorongan, dukungan, masukan, pengertian, kesabaran, kasih sayang dan dukungan doa selama ini kepada penulis.
3.
Sandi Prayitno, terima kasih atas segala dukungan, kesabaran, kasih sayang, perhatian, masukan, doa dan support selama ini kepada penulis.
4.
Bapak Dr. Dharmaputra Palekahelu, M.Pd selaku Dekan Fakultas Teknologi Informasi Universitas Kristen Satya Wacana Salatiga.
iv
5.
Bapak Prof. Dr. Ir. Eko Sediyono, M.Kom selaku Kaprogdi Magister Sistem Informasi Universitas Kristen Satya Wacana Salatiga yang telah memberikan bantuan dan dorongan semangat kepada penulis.
6.
Bapak Prof. Ir. Danny Manongga, M.Sc., Ph.D atas kesediaannya menjadi dosen pembimbing pertama, yang telah sabar dalam memberikan bimbingan dan pengarahan selama penyusunan thesis ini.
7.
Bapak Dr. Ir. Wiranto Herry Utomo atas kesediaannya menjadi dosen pembimbing kedua, yang telah sabar dalam memberikan bimbingan, masukan dan pengarahan selama penyusunan thesis ini.
8.
Semua dosen dan staff di Fakultas Teknologi Informasi, Program Studi Magister Sistem Informasi UKSW yang tidak dapat disebutkan satu persatu, thanks for your helps
9.
Temen-temen seperjuangan MSI angkatan 08 (kak bonie, kak rita, marlin, widi, pak andi, daniel, arao, astriyer), serta kakak dan adik angkatan MSI yang telah banyak membantu penulis dalam menyelesaikan studi S2. I Love you all!!
10.
Teman-teman di Fakultas Teknologi Informasi UKSW yang tidak dapat disebutkan satu persatu, Terimakasih buat persahabatan hingga akhir kuliah, semoga kalian semua tetap kompak, seru, rukun, oce!!!
11.
Teman-teman ngajar, teman-teman kos, teman-teman dolan, dan yang tidak dapat disebutkan satu persatu. Terima kasih sudah menjadi temanku yang rame, gila, kompak, dan baik.
v
12.
Teman-teman di UKSW dan semua pihak yang tidak dapat penulis sebutkan satu persatu hingga selesainya thesis ini, terimakasih all. SEMANGAT Semua.... Jiao You ya!!!...
13.
Murid-murid dan anak-anak mahasiswa, terimakasih atas kekritisan
kalian
dalam
mengkuti
kelas
dan
telah
menghadirkan keceriaan dalam kelas. Sukses untuk kalian semua yaaa..... 14.
Ke-empat anjing ku: Balto, Snoppy, Milo, Chelsea yang luculucu ’n imut-imut yang membantu penulis dalam melepas stress, penat, dan lelah dalam pengerjaan thesis. Penulis menyadari bahwa Thesis ini masih jauh dari
kesempurnaan, namun demikian penulis berharap semoga dapat bermanfaat bagi semua pembaca. Terima kasih, Tuhan berkati.
Salatiga, Juni 2013
(Isadora Nugroho)
vi
Daftar Isi Halaman Judul .................................................................................. Halaman Pengesahan ........................................................................ Halaman Pernyataan ......................................................................... Prakata ............................................................................................ Daftar Isi ........................................................................................... Daftar Gambar .................................................................................. Daftar Tabel ..................................................................................... Daftar Lampiran ................................................................................ Daftar Singkatan ............................................................................... Abstract ............................................................................................ Bab 1 Introduction ......................................................................... Bab 2 Literature Review ................................................................ Bab 3 Customer Analysis ............................................................... 3.1 Customer Loyalty .......................................................... 3.2 Service Quality .............................................................. 3.3 Relationship between Customer Loyalty and Service Quality ............................................................................ Bab 4 ID3 Algorithm ..................................................................... 4.1 Entropy .......................................................................... 4.2 Information Gain............................................................ Bab 5 Data Preparation and Architecture Model ............................ Bab 6 Discussion ........................................................................... 6.1 Servqual Model Formulas .............................................. 6.2 Correlation Test between Variable ................................. 6.3 ID3 Algorithm Integration (Entropy & Information Gain) 6.3.1 Reliability Dimension............................................. 6.3.2 Ressponsiveness Dimension ................................... 6.3.3 Assurance Dimension ............................................. 6.3.4 Emphaty Dimension ............................................... 6.3.5 Tangibles Dimension.............................................. 6.4 Decision Tree................................................................. 6.5 Evaluation...................................................................... Bab 7 Analysis ............................................................................... Bab 8 Conclusion........................................................................... References ........................................................................................ Lampiran ........................................................................................
vii
Hal i ii iii iv vii viii ix x xi 1 1 1 1 1 2 2 2 3 3 3 3 4 4 4 5 5 5 5 5 5 5 6 7 7 8
Daftar Gambar Hal Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8
Correlation between Service Quality and Customer Loyalty ......................................................................... Architectural Model Flowchart ..................................... Servqual Method with ID3 ............................................ Correlation Test between Variable ................................ Target Attribute ............................................................ Servqual Attribute Dimension ....................................... Tangibles Attribute ....................................................... Loyal Customer Factor..................................................
viii
2 3 3 4 4 5 5 6
Daftar Tabel Table 1 Table 2
Variable in Servqual Model .......................................... ID3 Algorithm Result Table ..........................................
ix
Hal 4 6
Daftar Lampiran Kuesioner Penelitian.......................................................................... Dimensi Kehandalan (Reliability) ...................................................... Dimensi Ketanggapan (Responsiveness)............................................ Dimensi Jaminan Mutu (Assurance) .................................................. Dimensi Empati (Emphaty) ............................................................... Dimensi Fisik (Tangible) ...................................................................
x
Hal 8 9 9 10 10 11
Daftar Singkatan ID3 Servqual
: Iterative Dichotomizer Tree : Service Quality
xi
International Journal of Computer Applications (0975 – 8887) Volume 69– No.11, Mey 2013
ID3 Algorithm to Identify Customer Loyalty Factor at Semarang Ceramics Company Isadora Nugroho
Danny Manongga
Wiranto H. Utomo
Master of Information Systems, Faculty of Information Technology, Satya Wacana Christian University, Indonesia
Information System Department Faculty of Information Technology, Satya Wacana Christian University, Indonesia
Information System Department Faculty of Information Technology, Satya Wacana Christian University, Indonesia
[email protected]
[email protected]
[email protected]
Performance Evaluation of Decision Tree Classifiers on Medical Datasets [1]. This research is a broad material, so the researcher decide to use decision tree to diagnose disease such as Brest cancer diagnose with ultrasonic picture, ovaries cancer and heart beat diagnose. The research compares analysis accuracy and time complexity by using ID3, C4.5 and CART. These tools is able to read and translate to medical diagnose. The result of this research is decision tree works well in the medical classification data.
ABSTRACT The aim of this research is to identify the customer loyalty at a Semarang ceramics company. The research process uses ID 3 algorithm and 5 SERVQUAL attributes, reliability, assurance, tangible, empathy, and responsiveness. Questioner data is the main data, which is analyzed by WEKA 3.7.7 software. The result of this research is Responsiveness attribute and its indicator quick service to the customer is the main factor which is influence the customer loyalty.
The last research is Case Study on High Dimensional Data Analysis Using Decision Tree Model [12]. This research is to predict the disease possibility in a region. This research, which identificate the parameter significant in prediction process, uses Decision tree. There are 3 models which is created trough ID3 Algorithm, unsupervised model to identification low level family characteristics and supervised model to high level family. Grouping model is enabling to understand the disease history group and decision tree is used to solve problem. The result is risk factor like climate, rainfall, deadly disease spread, water, temperature and environment is the highest factor which influences the low level family life and hereditary health history is for high level one.
Keywords Decision Tree, Algorithm ID3, Service Quality, Customer Loyalty
1. INTRODUCTION Recently, ceramics import from other country enters Indonesia and they give the best offer to the customer. Mirrored to this situation, local company should prepare and plan some strategic in providing the best service to the customer. With good strategic, local company can compete to foreign ceramics company especially to reach the customer loyalty. This research observes the problem in Semarang Ceramics Company, Indonesia. There are some factors to reach customer loyalty, first, creating a good relationship to the customer, give the best service and maintain the customer satisfaction. To implement these three factors is not as easy as turning a hand. Company should exert every effort to reach the customer satisfaction and profit company so the company can survive in the middle of competition. Company should prepare and plan the best strategy to face the competition and to reach the potential customer. There are 5 dimension which influence the customer loyalty, reliability, assurance, tangible, empathy, and responsiveness. In measuring customer satisfaction and customer loyalty, analyzed and data observed should be held to find the factors of the problem through quisioner method. Some journals use ID3 Algorithm as decision tree in every problem. ID3 concept implementation could help the company to identify potential customer. From the tools, the company can understand the customer behavior as a recommendation to marketing implementation strategy which can reach the best benefit to Semarang Ceramics Company.
In calculating and creating decision tree, this research uses ID3 Algorithm. Customer questionnaire result, with 5 dimensions in SERVQUAL, is the research data. In calculation process, the system is automatically calculating in ID3 Algorithm. The result of this research is loyalty customer format, loyal and non-loyal customer to the company.
3. CUSTOMER ANALYSIS This research focuses on customer analysis especially customer behavior. The aim is to identify the loyalty customer factors and its relation to customer service quality of the company. These factors are divided to some variable and attribute which is the most influence to the company. From the variable data, it can conclude basic factor of customer loyalty to the company.
3.1 Customer Loyalty
2. LITERATURE REVIEW
Customer loyalty is repurchasing of a product or a service because of the interest of the product or the service. Customer loyalty is a continue manifest of customer satisfaction in using a product or service. There are some loyal customers like below [6]: a. Regularly purchase in the company. A loyal customer repurchases repeatedly in the same company in a period.
There are some previous researches related to the theory. The first title is ID3 Classifier for Pupils’ Status Prediction [8]. The research explained that decision tree is used as a basic to predict required student status to further studies. Data mining is the source data of this research. The result of this research is the student classification which has high academic output. Second research evaluated the patient disease diagnose,
1
International Journal of Computer Applications (0975 – 8887) Volume 69– No.11, Mey 2013 b. Purchase out of common product/service. Loyal customer doesn’t only purchase a product, but also purchase another product or service in the same company. c. Give a recommendation to others. Loyal customer will recommend to others about their positive experience in purchasing a product or a service in a company. They persuade other not to purchase in the other company. d. Show uninterested to try another product. They trust the product or a service, that they use is the best product and different to other.
3.2 Service Quality A group of attribute is needed to analyze service quality. Attribute is behavior or specific characteristics which represent and point of the service [4]. Service dimension is the main thought in deciding the characteristics of the service. There are 5 dimensions to analyze the service:
Figure 1: Correlation between Service quality and Customer loyalty
1. Tangible Attribute and dimension, which include physical facilities, to fulfill the customer satisfaction, like: building, equipment. 2. Reliability Attribute and dimension, which include capability in servicing to the customer like promised. 3. Responsiveness Willingness to help customer and responsive to what customer need. Capability to give proper information and always give the best service as soon as possible. 4. Assurance Make a safety feeling and comfortable to the customer. These conditions influence to the customer because of its credibility, politeness, hospitality and employee knowledge in servicing customer. 5. Emphaty Give more attention to the customer. Simplicity of making a relation, good communication, personal attention and understand the customer needs.
4. ID3 ALGORITHM One of induction decision tree algorithms is ID3 (Iterative Dichotomies 3). ID3 built by J. Ross Quinlan [1]. ID3 Algorithm could be implemented to Recursive function (function which can call it). ID3 is basic decision tree learning algorithm. Algorithm searches greedily in all decision tree possibilities. ID3 characteristics in bulding decision tree are from root to leaf (top-down) and training data recursively. It is parted to a small categories in built the tree (divide and conquer). The first attribute should be plant on the root, and then it is evaluated by statistics information gain measurement. Data group consist of decision variable and outcome variable. Both variables should have categorical value. ID3 need the value as a label and symbol which doesn’t have relation one another. Below is the ID3 Algorithm:
3.3 Relation between Service Quality and Customer Loyalty
1. Tree is started as a single root, which represent all data.
Service quality and customer loyalty has a tight correlation. If the company can fill customer’s expectation, need and wish, the value of the service quality is good. They will be a loyal customer if their need, wish and expectation about the product are filled. For sure, they will come back and repurchase a product or service in the same company and don’t have any interest to purchase another product from competitor. Having loyal customer is the end of the business. To reach the aim, the company should improve their service quality with focus on 5 dimensions, tangible, reliability, responsiveness, assurance, and empathy. Through the efforts, the company can fill customer need, hope and expectation. It can attract loyal customer to come to purchase in the company, like shown in Figure 1.
2. After node root, the data will measured by information gain to select which attribute can be divided attribute. 3. A branch of the tree is made from the divided attribute and the data will distribute to other branch. 4. Algorithm will use the same process (recursive) to make secession tree. When the attribute is the dividing node or branch, the attribute will be kept out from information gain value. 5. Recursive dividing process will stop if one of the conditions is not fulfilled:
2
All data from branch include in the same class.
All attribute is already used, but the rest is in the different class. In this case, we should take the data which represent the most class to be class label in the node leaf.
International Journal of Computer Applications (0975 – 8887) Volume 69– No.11, Mey 2013
There is no data on the new branch. In this case, we should take from the previous node and the data is taken which represent the most class to be class label.
Start
Identification Servqual 5 variables
4.1 Entropy A classified object on the tree should be tested entropy value. Entropy is a measurement of the information theory to get the characteristics of impurity and homogeneity from the group data. From entropy value, we can count the information gain value (IG) from each attribute [8].
Check Completeness Attributes
Count Value Entropy
Entropy(S) = -pa log2 pa - pb log2 S : Sample Room (data) to train. Pa : Positive Solution amount (supported) of the sample data in some criteria. Pb : Negative Solution (unsupported) of the sample data in some criteria.
Count Value Gain NO YES
From the entropy formula show that Entropy (S) is a needed bit to be extracted in a class (a or b) from the random data in an S sample room. It can conclude that the smaller entropy value, the better extraction in a class. The code, which gives optimal information, is –log2p bits. It is to inform probability p. Bit amount to extract S in to a class is -pa log2 pa - pb log2.
4.2 Information Gain
Gain the Highest Value
Check the rest of Attributes
After getting entropy value in a group data, we can measure the attribute affectivity to classify the data. It is information gain. Information gain of an attribute is like bellow [8]:
Result
End
A : attribute v : a possibility value for an attribute Values(A) : a possibility assemblage for an attribute |Sv| : Sample amount for v value |S| : Total sample data Entropy(Sv) : entropy for sample v value
5. DATA PREPARATION AND ARCHITECTURE MODEL The method of collecting data is questionnaire method. The questionnaires spread out to the customer, which is direct purchase and online purchase in a Semarang ceramics company. There are 128 data collected and will be analyzed. The analysis is to solve the problem which shows up in the company, especially Service quality dimension. Flowchart model as shown in Figure 2.
Figure 2: Architectural Model Flowchart
6. DISCUSSION The analysis method to implement ID3 Algorithm based on the scope of company service quality especially in Semarang Ceramics Company to the customer. Qualitative approach is used to make quantitative approach. Case study of this research is the service quality implementation can be seen in Figure 3. Servqual Model Formulas
Test the Correlation Between Variable
Integration Algorithm ID 3 (Entropy & Information Gain )
Decision Tree
Evaluation
Figure 3: Servqual Method with ID3
International Journal of Computer Applications (0975 – 8887) Volume 69– No.11, Mey 2013
6.1 Servqual Model Formulas Model Servqual variable is like below: 1. Tangibles (X1) surface of service quality like physical facilities, equipment, personnel and communication. Assurance (X4)
2. Reliability (X2) Capability to show the best service like promises punctual and trustable. Service should be punctual and in time without any fault. 3. Responsiveness (X3) willingness to help the customer and give a proper service. Long service cause negative response about the company service. 4. Assurance (X4) knowledge, politeness and capability to say trust and assure to the customer and give safety feeling or guarantee to the service.
Emphaty (X5)
5. Empathy (X5) give attention and understanding to customer personally as shown in Table 1. Table 1. Variable in Servqual Model Variable Tangibles (X1)
Reliability (X2)
Responsiveness (X3)
Operational Real aspect of the physical appearance.
Capability to serve on time, punctual and trustable without any mistake
Willingness to help customer and give a proper service.
Indicators - Product - Modern equipment - Facilities - Employee Neat - Consistent - On time - Serve like promised - Care - Sincerity - Clear Information - Service Speed - Service
Knowledge, politeness and employee capability to say trust and say trust to the customer. Give safety feeling or guarantee. Give a personal attention to the customer.
accuracy - Willingness to help - Readiness to respond - Trust - Safe - Comfortable - Polite - Guarantee
- Personal Attention - Proper operational time - Specifically need understanding - Work time suitability
In the research, service quality variable will represented trough alternative respond like: Good, Bad, Quick, Slow, Convince, Satisfaction, Comfortable and Less.
6.2 Correlation Test between Variable If attribute and variable are correlated, the relation can be shown in Figure 4.
Figure 4: Correlation Test between Variable
6.3 ID3 Algorithm Integration (Entropy & Information Gain) Target Attribute is customer loyalty which has loyal and disloyal value, while dimension is Reliability, Responsiveness, Assurance, Empathy, and Tangibles. There are 128 services from customer types, and the loyalty can be shown in Figure 5.
Figure 5: Target Attribute There are 90 types of loyal customer and 30 types of disloyal customer. The visualization can be shown in Figure 6.
4
International Journal of Computer Applications (0975 – 8887) Volume 69– No.11, Mey 2013
6.3.2 Responsiveness Dimension There are 50 quick value variables from loyal type, 13 quick variable from disloyal type, 15 slow value variable from loyal type and 50 slow value variables from disloyal.
6.3.3 Assurance Dimension There are 29 low value variables from loyal type, 12 low variables from disloyal, 32 convince value variable from loyal type and 55 convince value variable from disloyal.
6.3.4 Emphaty Dimension There are 61 satisfy value variables from loyal type, 25 satisfy value variables from disloyal, 12 less satisfy value variable from loyal type and 30 less satisfy value variable from disloyal.
Figure 6: Servqual Attribute Dimension
6.3.1 Reliability Dimension There are 64 well worth variable from loyal type. 33 well worth variables for disloyal type, 18 variables with a bad value for loyal type and 13 variables with a bad value for disloyal type.
6.3.5 Tangibles Dimension There are 21 less comfortable value variables from loyal type, 22 less comfortable value variables from disloyal, 29 comfortable value variable from loyal type and 30 comfortable value variable from disloyal.
6.4 Decision Tree
Figure 6: Tangibles Attribute Figure 6 is a tracking decision tree problem trough ID3 Algorithm. It shows that Responsiveness is a root of the tree. The tree has some leaf, Reliability, Empathy, Assurance. Empathy, Reliability, Tangibles, Assurance. Empathy, Reliability, Assurance. Empathy, Assurance, Tangibles, Reliability.
6.5 Evaluation From the decision tree, there is ID3 algorithm which shows the customer loyalty pattern from Service quality point of view. Some factors which cause the loyal customer can be seen from Figure 7.
5
International Journal of Computer Applications (0975 – 8887) Volume 69– No.11, Mey 2013
Ressponsiveness Quick Reliability=good Reliability=bad, Emphaty=satisfy Reliability=bad, Emphaty=less satisfy, Assurance= convince Slow Emphaty=satisfy, Reliability=good, Tangibles=less comfortable, Assurance= convince Emphaty=satisfy, Reliability=good, Tangibles=comfortable Emphaty=satisfy, Reliability=bad, Assurance=convince Emphaty=less satisfy, Assurance=convince, Tangibles=comfortable, Reliability=good Figure 7: Loyal Customer Factor
7. ANALYSIS From the research of service quality uses Servqual method in ID3 Algorithm; show that loyal customer can be shown in the Table 2. Table 2. ID3 Algorithm result table Reliability
Responsiveness
Assurance
Emphaty
Tangibles
Good
Quick
Convince
Satisfy
Comfortable
Good
Quick
Convince
Satisfy
Less Comfortable
Good
Quick
Convince
Less Satisfy
Comfortable
Good
Quick
Convince
Kurang
Less Comfortable
Good
Quick
Less Convince
Satisfy
Comfortable
Good
Quick
Less Convince
Satisfy
Less Comfortable
Good
Quick
Less Convince
Kurang
Comfortable
Good
Quick
Less Convince
Kurang
Less Comfortable
Good
Slow
Convince
Satisfy
Comfortable
Good
Slow
Convince
Satisfy
Less Comfortable
Good
Slow
Convince
Kurang
Comfortable
Good
Slow
Less Convince
Satisfy
Comfortable
Bad
Quick
Convince
Satisfy
Comfortable
Bad
Quick
Convince
Satisfy
Less Comfortable
Bad
Quick
Convince
Kurang
Comfortable
Bad
Quick
Convince
Kurang
Less Comfortable
Bad
Quick
Less Convince
Satisfy
Comfortable
Bad
Quick
Less Convince
Satisfy
Less Comfortable
Bad
Slow
Convince
Satisfy
Comfortable
Bad
Slow
Convince
Satisfy
Less Comfortable
From the all the variable Reliability, Responsiveness, Assurance, Empathy, and Tangibles can be conclude that: If they have quick responses and good reliability, the other variable will not be influenced. But if they has bad reliability, the other factor that should be focus is Empathy, should satisfy the customer. If the empathy is not satisfying enough, Assurance factor will play the main role and should convince the customer.
If the response is slow, the other factor which should be focus on is empathy and Assurance or empathy and Tangibles. Empathy should satisfy the customer. Assurance should assure the customer and tangibles should make the customer feel comfortable in purchasing or transacting. While Reliability doesn’t have a big influence if the responsiveness to the customer is slow.
6
International Journal of Computer Applications (0975 – 8887) Volume 69– No.11, Mey 2013
[6] H. S. Soliman, "Customer Relationship Management and
8. CONCLUSION
Its Relationship to the Marketing Performance", International Journal of Business and Social Science, Vol. 2 No. 10, June 2011.
Decision tree in ID3 Algorithm and WEKA 3.7.7 can solve some problems in understanding the customer loyalty factor. Both are easy to understand especially to analyze the loyalty factor. The main factor of this analysis is responsiveness. It is related to the employee capability to help the customer and give a quick response to the customer. Responsiveness is a dynamic dimension. It is influence the employee behavior, like Clear Information, Service Speed, Service accuracy, Willingness to help, and Readiness to respond.
[7] J. Munusamy, S. Chelliah, and H. W. Mun, "Service Quality Delivery and Its Impact on Customer Satisfaction in the Banking Sector in Malaysia", International Journal of Innovation Management and Technology, Vol. 1, No. 4, October 2010, ISSN: 20100248.
[8] K. Nandhini, and S. Saranya, "ID3 Classifier for Pupils'
9. REFERENCES [1] D. Lavanya, and K. U. Rani, "Performance Evaluation of
Status Prediction", (IJCA) International Journal of Computer Applications, Vol. 57, November 2011.
Decision Tree Classifiers on Medical Datasets", (IJCA) International Journal of Computer Applications, Vol. 26 No. 4, July 2011.
[9] Nino, 2012, Service Quality Infulence Analysis and Relationship Quality to Customer Loyalty on AHASS Tamtama Motor in Kediri Region, REVITALISATION Management Journal, Vol.1, No.2, September 2012.
[2] D. L. Gupta, A. K. Malviya, and S. Singh, "Performance Analysis of Classification Tree Learning Algorithms", (IJCA) International Journal of Computer Applications, Vol. 55 No. 6, October 2012.
[10] N. Malpathak, S. Shinde, and R.George, "Cloud Based CRM Application", (IJCA) International Journal of Computer Applications, April 2012, ISSN: 0975-8887.
[3] D. S. Kumar, N. Ananthi, and M. Devi, "An Approach to
[11] R. A. Soeini, B. Jafari, and M. Abdollahzadeh, "CRM
Automation Selection of Decision Tree based on Training Data Set", (IJCA) International Journal of Computer Applications, Vol. 64 No. 21, February 2013.
Performance Measurement Process", (IJES) International Journal of Emerging Sciences, March 2012, ISSN: 22224254.
[4] Firdian Endy, Surachman, and Santoso Budi, 2012,
[12] T. Smitha, and V. Sundaram, "Case Study On High
Servqual Application Method and Six Sigma to Analise PT. PLN (Persero) Service Quality Network Service Unit (UPJ) Dinoyo Malang, Science and Engineering, Vol.13 No.3 September 2012, ISSN: 1412-7237.
Dimensional Data Analysis Using Decision Tree Model", (IJCSI) International Journal of Computer Science Issues, Vol. 9, Issue 3, No. 3, May 2012, ISSN: 16940814.
[5] H. Monem, A. R. C. Hussin, and R. Sharifian, "Comprehensive Prioritized Perspectives of CRM System for Implementation in the Developing Countries Hospitals", (IJCA) International Journal of Computer Applications, Vol. 50 No. 20, July 2012.
7
Lampiran KUESIONER PENELITIAN Pengantar Saya mengucapkan terima kasih atas perhatian dan respon yang Bapak/Ibu berikan dengan bersedia mengisi kuesioner ini. Kuesioner ini diedarkan dalam rangka penelitian yang sedang dilakukan untuk menguji keberhasilan pelayanan PT SCI terhadap customernya. Sehubungan dengan itu, Bapak/Ibu diminta untuk membaca dengan teliti pertanyaan dan pernyataan di bawah ini. Kemudian dimohon untuk menjawab pertanyaan dan pernyataan yang ada dengan mengisi jawaban tertulis untuk jawaban yang bersifat terbuka pada kotak pilihan jawaban yang tersedia dengan sejujurnya sesuai dengan kondisi yang Bapak/Ibu rasakan. Tidak ada jawaban benar atau salah. Setiap jawaban yang Bapak/Ibu berikan akan saya apresiasi. Saya akan menjaga kerahasiaan jawaban yang Bapak/Ibu berikan. Demikianlah permohonan saya, terima kasih atas kerjasama dan waktu yang telah Bapak/Ibu luangkan untuk mengisi kuesioner ini. Profil Responden Jawaban diisi di tempat yang telah disediakan, untuk jawaban pilihan mohon dilingkari satu jawaban yang benar. Nama : (Boleh tidak diisi) Jenis Kelamin : L / P Status : Menikah / Belum Menikah Usia : tahun 1. Tingkat Pendapatan a. <= 5juta b. > 5 juta s/d 10 juta c. > 10 juta s/d 25 juta 8
d. > 25 juta s/d 50 juta e. > 50 juta s/d 100 juta f. > 100 juta 2. Sudah berapa lama anda menjadi Customer PT SCI ? a. < 1 tahun b. 1 tahun s/d 2 tahun c. 2,1 tahun s/d 3 tahun d. > 4 tahun 3. Apakah anda membeli produk keramik di Perusahaan lain selain di PT SCI? a. Ya b. Tidak 4. Apa alasan anda memilih Perusahaan lain selain PT SCI ? a. Keamanan b. Fasilitas c. Pelayanan (Service) d. Lainnya, sebutkan ______________________________________ Data ini akan ditangani dengan penuh kerahasiaan Berdasarkan pengalaman Bapak/Ibu sebagai sebagai customer dari PT SCI, kami memohon kesediaan anda untuk memikirkan tentang kualitas pelayanan perusahaan perbankan yang ideal menurut anda. Mohon tunjukkan pendapat anda dengan memberi tanda centang (√) pada jawaban yang Bapak/Ibu anggap paling mewakili keadaan Bapak/Ibu.
DIMENSI KEHANDALAN (RELIABILITY) Kemampuan karyawan untuk memberikan pelayanan dijanjikan dengan baik dan tepat. No.
Pernyataan
1.
Bagaimana pelayanan perusahaan, ketika anda memiliki masalah dan apakah pihak perusahaan cepat serta tanggap dalam menyelesaikannya Apakah pelayanan pihak perusahaan baik pertama kalinya Bagaimana reputasi pelayanan karyawan
2. 3.
9
Baik
yang Buruk
perusahaan di mata customer Sehingga secara umum anda menyimpulkan bahwa kemampuan karyawan untuk memberikan pelayanan yang dijanjikan dengan baik dan tepat adalah… DIMENSI KETANGGAPAN (RESPONSIVENESS) Ketersediaan Perusahaan untuk membantu customernya dan memberikan pelayanan yang cepat No.
Pernyataan
Cepat
Lambat
1.
Bagaimana karyawan perusahaan memberikan pelayanan yang tepat di saat customer membutuhkannya 2. Bagaimana reaksi karyawan perusahaan dalam memberikan pelayanan secara tepat kepada customer 3. Bagaimana perilaku karyawan dalam menunjukkan reaksi tidak pernah terlalu sibuk dan selalu memiliki waktu terhadap customer Sehingga secara umum anda menyimpulkan bahwa ketersediaan Perusahaan untuk membantu customernya dan memberikan pelayanan yang cepat adalah… DIMENSI JAMINAN MUTU (ASSURANCE) Pengetahuan dan keramahan dari karyawan perusahaan dan kemampuan untuk memberikan rasa aman dan percaya terhadap customer. No.
Pernyataan
1.
Bagaimana tanggapan karyawan saat menjawab pertanyaan-pertanyaan yang anda ajukan, apakah memiliki pengetahuan yang cukup. Bagaimana sikap karyawan perusahaan dalam membuat anda yakin dengan kualitas pelayanan yang mereka berikan.
2.
Meyakinkan
10
Kurang
3.
Bagaimana kekonsistenan karyawan dalam bersikap ramah terhadap anda Sehingga secara umum anda menyimpulkan bahwa pengetahuan dan keramahan dari karyawan perusahaan dan kemampuan untuk memberikan rasa aman dan percaya terhadap customer adalah… DIMENSI EMPATI (EMPATHY) Kepedulian perusahaan dalam memberikan perhatian secara khusus kepada customernya. No.
Pernyataan
Memuaskan
Kurang
1.
Bagaimana pelayanan perusahaan dalam memberikan perhatian khusus secara individu 2. Bagaimana operasi layanan dalam memberikan waktu yang sesuai kepada customernya 3. Bagaimana pola karyawan perusahaan dalam memahami kebutuhan khusus masing-masing customer Sehingga secara umum anda menyimpulkan bahwa kepedulian perusahaan dalam memberikan perhatian secara khusus kepada customernya adalah… DIMENSI FISIK (TANGIBLE) Penampilan dari fasilitas fisik (gedung, kursi, dll), perlengkapan, karyawan, dan alat- alat komunikasi. No.
Pernyataan
1.
Apakah fasilitas fisik (kursi, meja, gedung, dll) terlihat nyaman dan menarik. Apakah peralatan pelayanan customer seperti brosur, poster dan pamphlet terlihat jelas, nyaman, dan menarik
2.
11
Nyaman
Kurang
3.
Apakah karyawan berpenampilan rapi dan meja kerja tertata rapi dan nyaman. Sehingga secara umum anda menyimpulkan bahwa penampilan dari fasilitas fisik (gedung, kursi, dll), perlengkapan, karyawan, dan alatalat komunikasi. adalah…
12
Isadora Nugroho
IJCA May 2013 Edition: Letter of Acceptance 1 pesan Editor IJCA <[email protected]> 4 Mei 2013 19.08 Kepada: [email protected], [email protected], [email protected] Dear Isadora Nugroho, Danny Manongga and Wiranto H. Utomo, We are delighted to inform that your research paper has been “Accepted for Publication” in International Journal of Computer Applications (IJCA) May 2013 Edition. Paper Title: ID3 Algorithm to Identify Customer Loyalty Factor at
Semarang Ceramics Company Paper Reference ID: pxc3887900 The important deadlines regarding publication are as follows: Camera ready paper submission : May 15, 2013 Copyright transfer : May 15, 2013 (IJCA Copyright Form) Publication fees transfer : May 15, 2013 The research article will be published in IJCA Digital Library on May 17, 2013. All the selected papers will be published in IJCA Digital Library with initial indexing with Google Scholar, CiteSeer, UlrichsWeb and ScientificCommons Index, University of St. Gallen, Switzerland. The articles are also indexed with SAO/NASA ADS Physics Abstract Service supported by Harvard University and NASA, Informatics and ProQuest CSA Technology Research Database. International Journal of Computer Applications is a voting member of CrossRef, USA and each of its research articles is allotted a unique DOI reference. Article Processing Fees
Processing charge for each general paper : USD 62.50 (INR 2900)
The Charges includes publication, indexing, maintenance of link resolvers and journal infrastructures. Find attached IJCA Payment Instructions for the payment transfer guidelines and the mode of payment according to your domicile country. The invoice/ receipt will be delivered to you on subject to realization of the payment. You are requested to re-submit your paper as Camera Ready Copy (CRC) in DOC/ DOCX and/or Latex file format conforming to the
13
prescribed IJCA Paper Template on or before May 15, 2013. Also find attached the copy of the Review Report upon your research article along with the IJCA template. IMPORTANT: Conformance to IJCA template is mandatory for inclusion and citation with Google Scholar, CiteSeer, UlrichsWeb, NASA ADS and ProQuest CSA Technology Research Database. Thank you for your association with IJCA. Best regards, Editorial Support Team, International Journal of Computer Applications, Foundation of Computer Science, New York, USA. www.ijcaonline.org
[ Call for Paper (Open): http://www.ijcaonline.org/calls- ] 3 lampiran IJCA Paper Template.doc 47K IJCA Payment Instructions.pdf 49K Review Report pxc3887900.doc
14