Customer Churn Prediction in Telecommunication A Decade Review and Classification (original) (raw)

Abstract

Acquisition and the retention of customers are the top most concerns in today's business world. The rapid increase of market in every business is leading to higher subscriber base. Consequently, companies have realized the importance of retaining the on hand customers. It has become mandatory for the service providers to reduce churn rate because the negligence could be resulted as profitability reduction in major perspective. Churn prediction helps in identifying those customers who are likely to leave a company. Telecommunication is coping with the issue of ever increasing churn rate. Data mining techniques enable these telecommunication companies to be equipped with effective methods for reducing churn rate. The paper reviews 61 journal articles to survey the pros and cons of renowned data mining techniques used to build predictive customer churn models in the field of telecommunication and thus providing a roadmap to researchers for knowledge accumulation about data mining techniques in telecom.

Figures (8)

Fig. 1 : Classification by Dataset Type Table 1. is showing the statistics that 4.92% (3 among 61 articles) research has been done on taking fixed line telecommunication data type. And 57.38% (35 among 61) papers worked upon wireless or mobile telephony churn prediction.

Fig. 1 : Classification by Dataset Type Table 1. is showing the statistics that 4.92% (3 among 61 articles) research has been done on taking fixed line telecommunication data type. And 57.38% (35 among 61) papers worked upon wireless or mobile telephony churn prediction.

Table 2: Classification by Techniques 4.2 Distribution of articles by Techniques Table 3 shows the distribution of journal articles by the technique being used for model building. Total 104 types of techniques have been used in 61 articles. Here it is a point to be noted that one article may have used more than single technique for model building.

Table 2: Classification by Techniques 4.2 Distribution of articles by Techniques Table 3 shows the distribution of journal articles by the technique being used for model building. Total 104 types of techniques have been used in 61 articles. Here it is a point to be noted that one article may have used more than single technique for model building.

Fig. 3 : Classification by Two Dataset Type Neural Network: Currently, neural networks are used by researchers in the field of classification, clustering and prediction (Berry & Linoff, 2004; Turban et al., 2007). When applying neural network technique, it converts the data into dimensional array of neurons in an orderly manner which ultimately forms a prediction hierarch (Tsai and Lu, 2009; Song, Kim, Cho, & Kim, 2004 ). limensional array of neurons in an orderly manner Regression: Regression is also very popular technique for predicting behavior. It determines the impact of many independent variables to predict the possible reliance of one or more than one dependent variables (Bingquan Huang et al., 2012; Michel Ballings, Dirk Van den Poel, 2012; Marcin Owczarczuk, 2010; Jiayin Qi et al., 2006).

Fig. 3 : Classification by Two Dataset Type Neural Network: Currently, neural networks are used by researchers in the field of classification, clustering and prediction (Berry & Linoff, 2004; Turban et al., 2007). When applying neural network technique, it converts the data into dimensional array of neurons in an orderly manner which ultimately forms a prediction hierarch (Tsai and Lu, 2009; Song, Kim, Cho, & Kim, 2004 ). limensional array of neurons in an orderly manner Regression: Regression is also very popular technique for predicting behavior. It determines the impact of many independent variables to predict the possible reliance of one or more than one dependent variables (Bingquan Huang et al., 2012; Michel Ballings, Dirk Van den Poel, 2012; Marcin Owczarczuk, 2010; Jiayin Qi et al., 2006).

Table 3: Classification by Publication Y ear

Table 3: Classification by Publication Y ear

Table 4 shows the distribution of articles by journal. Total 32 journals were explored to find articles related to chum _ prediction § in telecommunication sector. "Expert System with Applications" leads the race with 16 significant articles. "Telecommunication Policy" is following the race with 6 papers published in context of customer churn prediction.

Table 4 shows the distribution of articles by journal. Total 32 journals were explored to find articles related to chum _ prediction § in telecommunication sector. "Expert System with Applications" leads the race with 16 significant articles. "Telecommunication Policy" is following the race with 6 papers published in context of customer churn prediction.

Fig. 4: Classification by mostly used techniques

Fig. 4: Classification by mostly used techniques

Loading...

Loading Preview

Sorry, preview is currently unavailable. You can download the paper by clicking the button above.

References (17)

  1. N.Kamalraj, and A.Malathi, "A Survey on Churn Prediction Techniques in Communication Sector", International Journal of Computer Applications, Volume 64, Isssue 5, 2013, pp. 39-42.
  2. Yen-Hsien Lee, Chih-Ping Wei, Tsang-Hsiang Cheng, Ching-Ting Yang, "Nearest-neighbor-based approach to time-series classification", Decision Support Systems, Volume 53, Issue 1, 2013, pp. 207-217.
  3. Chitra Phadke, Huseyin Uzunalioglu, Veena B. Mendiratta , Dan Kushnir , and Derek Doran , "Prediction of Subscriber Churn Using Social Network Analysis", Bell Labs Technical Journal, Volume 17, Issue 4, 2013, pages 63-75.
  4. Vivek Bhambri, "Data Mining as a Tool to Predict Churn Behaviour of Customers", GE-International Journal of Management Research (IJMR), April 2013, pages 59-69.
  5. N.Kamalraj and A.Malathi, "A Survey on Churn Prediction Techniques in Communication Sector", International Journal of Computer Applications, Volume 64-No.5, 2013, pp. 0975 -8887.
  6. Clement Kirui, Li Hong, Wilson Cheruiyot and Hillary Kirui, "Predicting Customer Churn in Mobile Telephony Industry Using Probabilistic Classifiers in Data Mining", International Journal of Computer Science Issues (IJCSI), Vol. 10, Issue 2, 2013, No 1, 165-172.
  7. Golshan Mohammadi,
  8. J Lu., "Modeling Customer Lifetime Value Using Survival Analysis: An Application in the Telecommunications Industry", Data Mining Techniques, 2003, SUGI 28.
  9. Kim, Y. H., and Moon, B. R., "Multicampaign assignment problem. Knowledge and Data Engineering", IEEE Transactions, Volume:18 , Issue: 3, 2003, 405-414.
  10. Rosset, S., Neumann, E., Eick, U., and Vatnik, N., "Customer lifetime value models for decision support", Data Mining and Knowledge Discovery ,Volume 7, Issue 3, 2003, pp 321-339
  11. Koh, H. C., and Chan, K. L. G., "Data mining and customer relationship marketing in the banking industry", Singapore Management Review, 24, 2002, 1-28.
  12. Chih Ping Wei, I-Tang Chiu, "Expert Systems with Applications", Volume 23, Issue 2, August 2002, Pages 103-112.
  13. Lejeune, M. A. P. M., "Measuring the impact of data mining on churn management", Internet Research: Electronic Networking Applications and Policy, 11, 2001, 375-387.
  14. Berson, A., Smith, S., and Thearling, K. (2000). Building data mining applications for CRM. McGraw-Hill.
  15. Madden, G., Savage, S. J., and Coble-Neal, G., "Subscriber churn in the Australian ISP market", Information Economics and Policy,11 (2), 1999, 195-207
  16. Nord, J. H., and Nord, G. D., "MIS research: Journal status and analysis", Information and Management, 1995, 29, 29-42.
  17. Berson, A., Smith, S., and Therling, K. (1999). Building data mining applications for CRM . New York: McGraw-Hill.